The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)

Today, we're joined by Shirley Wu, senior director of software engineering at Juniper Networks to discuss how machine learning and artificial intelligence are transforming network management. We explore various use cases where AI and ML are applied to enhance the quality, performance, and efficiency of networks across Juniper’s customers, including diagnosing cable degradation, proactive monitoring for coverage gaps, and real-time fault detection. We also dig into the complexities of integrating data science into networking, the trade-offs between traditional methods and ML-based solutions, the role of feature engineering and data in networking, the applicability of large language models, and Juniper’s approach to using smaller, specialized ML models to optimize speed, latency, and cost. Finally, Shirley shares some future directions for Juniper Mist such as proactive network testing and end-user self-service.

The complete show notes for this episode can be found at https://twimlai.com/go/710.

Today, we're joined by Jason Liu, freelance AI consultant, advisor, and creator of the Instructor library to discuss all things retrieval-augmented generation (RAG). We dig into the tactical and strategic challenges companies face with their RAG system, the different signs Jason looks for to identify looming problems, the issues he most commonly encounters, and the steps he takes to diagnose these issues. We also cover the significance of building out robust test datasets, data-driven experimentation, evaluation tools, and metrics for different use cases. We also touched on fine-tuning strategies for RAG systems, the effectiveness of different chunking strategies, the use of collaboration tools like Braintrust, and how future models will change the game. Lastly, we cover Jason’s interest in teaching others how to capitalize on their own AI experience via his AI consulting course.

The complete show notes for this episode can be found at https://twimlai.com/go/709.

Today we're joined by Sunil Mallya, CTO and co-founder of Flip AI. We discuss Flip’s incident debugging system for DevOps, which was built using a custom mixture of experts (MoE) large language model (LLM) trained on a novel "CoMELT" observability dataset which combines traditional MELT data—metrics, events, logs, and traces—with code to efficiently identify root failure causes in complex software systems. We discuss the challenges of integrating time-series data with LLMs and their multi-decoder architecture designed for this purpose. Sunil describes their system's agent-based design, focusing on clear roles and boundaries to ensure reliability. We examine their "chaos gym," a reinforcement learning environment used for testing and improving the system's robustness. Finally, we discuss the practical considerations of deploying such a system at scale in diverse environments and much more.

The complete show notes for this episode can be found at https://twimlai.com/go/708.

Today, we're joined by Scott Stephenson, co-founder and CEO of Deepgram to discuss voice AI agents. We explore the importance of perception, understanding, and interaction and how these key components work together in building intelligent AI voice agents. We discuss the role of multimodal LLMs as well as speech-to-text and text-to-speech models in building AI voice agents, and dig into the benefits and limitations of text-based approaches to voice interactions. We dig into what’s required to deliver real-time voice interactions and the promise of closed-loop, continuously improving, federated learning agents. Finally, Scott shares practical applications of AI voice agents at Deepgram and provides an overview of their newly released agent toolkit.

The complete show notes for this episode can be found at https://twimlai.com/go/707.

Today, we're joined by Tim Rocktäschel, senior staff research scientist at Google DeepMind, professor of Artificial Intelligence at University College London, and author of the recently published popular science book, “Artificial Intelligence: 10 Things You Should Know.” We dig into the attainability of artificial superintelligence and the path to achieving generalized superhuman capabilities across multiple domains. We discuss the importance of open-endedness in developing autonomous and self-improving systems, as well as the role of evolutionary approaches and algorithms. Additionally, we cover Tim’s recent research projects such as “Promptbreeder,” “Debating with More Persuasive LLMs Leads to More Truthful Answers,” and more.

The complete show notes for this episode can be found at https://twimlai.com/go/706.

Today, we're joined by Lucas García, principal product manager for deep learning at MathWorks to discuss incorporating ML models into safety-critical systems. We begin by exploring the critical role of verification and validation (V&V) in these applications. We review the popular V-model for engineering critical systems and then dig into the “W” adaptation that’s been proposed for incorporating ML models. Next, we discuss the complexities of applying deep learning neural networks in safety-critical applications using the aviation industry as an example, and talk through the importance of factors such as data quality, model stability, robustness, interpretability, and accuracy. We also explore formal verification methods, abstract transformer layers, transformer-based architectures, and the application of various software testing techniques. Lucas also introduces the field of constrained deep learning and convex neural networks and its benefits and trade-offs.

The complete show notes for this episode can be found at https://twimlai.com/go/705.

Today, we're joined by Arvind Narayanan, professor of Computer Science at Princeton University to discuss his recent works, AI Agents That Matter and AI Snake Oil. In “AI Agents That Matter”, we explore the range of agentic behaviors, the challenges in benchmarking agents, and the ‘capability and reliability gap’, which creates risks when deploying AI agents in real-world applications. We also discuss the importance of verifiers as a technique for safeguarding agent behavior. We then dig into the AI Snake Oil book, which uncovers examples of problematic and overhyped claims in AI. Arvind shares various use cases of failed applications of AI, outlines a taxonomy of AI risks, and shares his insights on AI’s catastrophic risks. Additionally, we also touched on different approaches to LLM-based reasoning, his views on tech policy and regulation, and his work on CORE-Bench, a benchmark designed to measure AI agents' accuracy in computational reproducibility tasks.

The complete show notes for this episode can be found at https://twimlai.com/go/704.

Today, we're joined by Shreya Shankar, a PhD student at UC Berkeley to discuss DocETL, a declarative system for building and optimizing LLM-powered data processing pipelines for large-scale and complex document analysis tasks. We explore how DocETL's optimizer architecture works, the intricacies of building agentic systems for data processing, the current landscape of benchmarks for data processing tasks, how these differ from reasoning-based benchmarks, and the need for robust evaluation methods for human-in-the-loop LLM workflows. Additionally, Shreya shares real-world applications of DocETL, the importance of effective validation prompts, and building robust and fault-tolerant agentic systems. Lastly, we cover the need for benchmarks tailored to LLM-powered data processing tasks and the future directions for DocETL.

The complete show notes for this episode can be found at https://twimlai.com/go/703.

Today, we're joined by Nicholas Carlini, research scientist at Google DeepMind to discuss adversarial machine learning and model security, focusing on his 2024 ICML best paper winner, “Stealing part of a production language model.” We dig into this work, which demonstrated the ability to successfully steal the last layer of production language models including ChatGPT and PaLM-2. Nicholas shares the current landscape of AI security research in the age of LLMs, the implications of model stealing, ethical concerns surrounding model privacy, how the attack works, and the significance of the embedding layer in language models. We also discuss the remediation strategies implemented by OpenAI and Google, and the future directions in the field of AI security. Plus, we also cover his other ICML 2024 best paper, “Position: Considerations for Differentially Private Learning with Large-Scale Public Pretraining,” which questions the use and promotion of differential privacy in conjunction with pre-trained models.

The complete show notes for this episode can be found at https://twimlai.com/go/702.

Today, we're joined by Simon Willison, independent researcher and creator of Datasette to discuss the many ways software developers and engineers can take advantage of large language models (LLMs) to boost their productivity. We dig into Simon’s own workflows and how he uses popular models like ChatGPT and Anthropic’s Claude to write and test hundreds of lines of code while out walking his dog. We review Simon’s favorite prompting and debugging techniques, his strategies for sidestepping the limitations of contemporary models, how he uses Claude’s Artifacts feature for rapid prototyping, his thoughts on the use and impact of vision models, the role he sees for open source models and local LLMs, and much more.

The complete show notes for this episode can be found at https://twimlai.com/go/701.

Today, we're joined by Shengran Hu, a PhD student at the University of British Columbia, to discuss Automated Design of Agentic Systems (ADAS), an approach focused on automatically creating agentic system designs. We explore the spectrum of agentic behaviors, the motivation for learning all aspects of agentic system design, the key components of the ADAS approach, and how it uses LLMs to design novel agent architectures in code. We also cover the iterative process of ADAS, its potential to shed light on the behavior of foundation models, the higher-level meta-behaviors that emerge in agentic systems, and how ADAS uncovers novel design patterns through emergent behaviors, particularly in complex tasks like the ARC challenge. Finally, we touch on the practical applications of ADAS and its potential use in system optimization for real-world tasks.

The complete show notes for this episode can be found at https://twimlai.com/go/700.

Today, we're joined by Peter van der Putten, director of the AI Lab at Pega and assistant professor of AI at Leiden University. We discuss the newly adopted European AI Act and the challenges of applying academic fairness metrics in real-world AI applications. We dig into the key ethical principles behind the Act, its broad definition of AI, and how it categorizes various AI risks. We also discuss the practical challenges of implementing fairness and bias metrics in real-world scenarios, and the importance of a risk-based approach in regulating AI systems. Finally, we cover how the EU AI Act might influence global practices, similar to the GDPR's effect on data privacy, and explore strategies for closing bias gaps in real-world automated decision-making.

The complete show notes for this episode can be found at https://twimlai.com/go/699.

Today, we're joined by Harrison Chase, co-founder and CEO of LangChain to discuss LLM frameworks, agentic systems, RAG, evaluation, and more. We dig into the elements of a modern LLM framework, including the most productive developer experiences and appropriate levels of abstraction. We dive into agents and agentic systems as well, covering the “spectrum of agenticness,” cognitive architectures, and real-world applications. We explore key challenges in deploying agentic systems, and the importance of agentic architectures as a means of communication in system design and operation. Additionally, we review evolving use cases for RAG, and the role of observability, testing, and evaluation tools in moving LLM applications from prototype to production. Lastly, Harrison shares his hot takes on prompting, multi-modal models, and more!

The complete show notes for this episode can be found at https://twimlai.com/go/698.

Today, we're joined by Siddhika Nevrekar, AI Hub head at Qualcomm Technologies, to discuss on-device AI and how to make it easier for developers to take advantage of device capabilities. We unpack the motivations for AI engineers to move model inference from the cloud to local devices, and explore the challenges associated with on-device AI. We dig into the role of hardware solutions, from powerful system-on-chips (SoC) to neural processors, the importance of collaboration between community runtimes like ONNX and TFLite and chip manufacturers, the unique challenges of IoT and autonomous vehicles, and the key metrics developers should focus on to ensure optimal on-device performance. Finally, Siddhika introduces Qualcomm's AI Hub, a platform developed to simplify the process of testing and optimizing AI models across different devices.

The complete show notes for this episode can be found at https://twimlai.com/go/697.

Today, we're joined by Ashley Edwards, a member of technical staff at Runway, to discuss Genie: Generative Interactive Environments, a system for creating ‘playable’ video environments for training deep reinforcement learning (RL) agents at scale in a completely unsupervised manner. We explore the motivations behind Genie, the challenges of data acquisition for RL, and Genie’s capability to learn world models from videos without explicit action data, enabling seamless interaction and frame prediction. Ashley walks us through Genie’s core components—the latent action model, video tokenizer, and dynamics model—and explains how these elements collaborate to predict future frames in video sequences. We discuss the model architecture, training strategies, benchmarks used, as well as the application of spatiotemporal transformers and the MaskGIT techniques used for efficient token prediction and representation. Finally, we touched on Genie’s practical implications, its comparison to other video generation models like “Sora,” and potential future directions in video generation and diffusion models.

The complete show notes for this episode can be found at https://twimlai.com/go/696.

Today, we're joined by Marius Memmel, a PhD student at the University of Washington, to discuss his research on sim-to-real transfer approaches for developing autonomous robotic agents in unstructured environments. Our conversation focuses on his recent ASID and URDFormer papers. We explore the complexities presented by real-world settings like a cluttered kitchen, data acquisition challenges for training robust models, the importance of simulation, and the challenge of bridging the sim2real gap in robotics. Marius introduces ASID, a framework designed to enable robots to autonomously generate and refine simulation models to improve sim-to-real transfer. We discuss the role of Fisher information as a metric for trajectory sensitivity to physical parameters and the importance of exploration and exploitation phases in robot learning. Additionally, we cover URDFormer, a transformer-based model that generates URDF documents for scene and object reconstruction to create realistic simulation environments.

The complete show notes for this episode can be found at https://twimlai.com/go/695.

Today, we're joined by Hamel Husain, founder of Parlance Labs, to discuss the ins and outs of building real-world products using large language models (LLMs). We kick things off discussing novel applications of LLMs and how to think about modern AI user experiences. We then dig into the key challenge faced by LLM developers—how to iterate from a snazzy demo or proof-of-concept to a working LLM-based application. We discuss the pros, cons, and role of fine-tuning LLMs and dig into when to use this technique. We cover the fine-tuning process, common pitfalls in evaluation—such as relying too heavily on generic tools and missing the nuances of specific use cases, open-source LLM fine-tuning tools like Axolotl, the use of LoRA adapters, and more. Hamel also shares insights on model optimization and inference frameworks and how developers should approach these tools. Finally, we dig into how to use systematic evaluation techniques to guide the improvement of your LLM application, the importance of data generation and curation, and the parallels to traditional software engineering practices.

The complete show notes for this episode can be found at https://twimlai.com/go/694.

Today, we're joined by Albert Gu, assistant professor at Carnegie Mellon University, to discuss his research on post-transformer architectures for multi-modal foundation models, with a focus on state-space models in general and Albert’s recent Mamba and Mamba-2 papers in particular. We dig into the efficiency of the attention mechanism and its limitations in handling high-resolution perceptual modalities, and the strengths and weaknesses of transformer architectures relative to alternatives for various tasks. We dig into the role of tokenization and patching in transformer pipelines, emphasizing how abstraction and semantic relationships between tokens underpin the model's effectiveness, and explore how this relates to the debate between handcrafted pipelines versus end-to-end architectures in machine learning. Additionally, we touch on the evolving landscape of hybrid models which incorporate elements of attention and state, the significance of state update mechanisms in model adaptability and learning efficiency, and the contribution and adoption of state-space models like Mamba and Mamba-2 in academia and industry. Lastly, Albert shares his vision for advancing foundation models across diverse modalities and applications.

The complete show notes for this episode can be found at https://twimlai.com/go/693.

Today, we're joined by Amir Bar, a PhD candidate at Tel Aviv University and UC Berkeley to discuss his research on visual-based learning, including his recent paper, “EgoPet: Egomotion and Interaction Data from an Animal’s Perspective.” Amir shares his research projects focused on self-supervised object detection and analogy reasoning for general computer vision tasks. We also discuss the current limitations of caption-based datasets in model training, the ‘learning problem’ in robotics, and the gap between the capabilities of animals and AI systems. Amir introduces ‘EgoPet,’ a dataset and benchmark tasks which allow motion and interaction data from an animal's perspective to be incorporated into machine learning models for robotic planning and proprioception. We explore the dataset collection process, comparisons with existing datasets and benchmark tasks, the findings on the model performance trained on EgoPet, and the potential of directly training robot policies that mimic animal behavior.

The complete show notes for this episode can be found at https://twimlai.com/go/692.

Today, we're joined by Sarah Bird, chief product officer of responsible AI at Microsoft. We discuss the testing and evaluation techniques Microsoft applies to ensure safe deployment and use of generative AI, large language models, and image generation. In our conversation, we explore the unique risks and challenges presented by generative AI, the balance between fairness and security concerns, the application of adaptive and layered defense strategies for rapid response to unforeseen AI behaviors, the importance of automated AI safety testing and evaluation alongside human judgment, and the implementation of red teaming and governance. Sarah also shares learnings from Microsoft's ‘Tay’ and ‘Bing Chat’ incidents along with her thoughts on the rapidly evolving GenAI landscape.

The complete show notes for this episode can be found at https://twimlai.com/go/691.

Today, we're joined by Eric Nguyen, PhD student at Stanford University. In our conversation, we explore his research on long context foundation models and their application to biology particularly Hyena, and its evolution into Hyena DNA and Evo models. We discuss Hyena, a convolutional-based language model developed to tackle the challenges posed by long context lengths in language modeling. We dig into the limitations of transformers in dealing with longer sequences, the motivation for using convolutional models over transformers, its model training and architecture, the role of FFT in computational optimizations, and model explainability in long-sequence convolutions. We also talked about Hyena DNA, a genomic foundation model pre-trained on 1 million tokens, designed to capture long-range dependencies in DNA sequences. Finally, Eric introduces Evo, a 7 billion parameter hybrid model integrating attention layers with Hyena DNA's convolutional framework. We cover generating and designing DNA with language models, hallucinations in DNA models, evaluation benchmarks, the trade-offs between state-of-the-art models, zero-shot versus a few-shot performance, and the exciting potential in areas like CRISPR-Cas gene editing.

The complete show notes for this episode can be found at https://twimlai.com/go/690.

Today, we're joined by Andres Ravinet, sustainability global black belt at Microsoft, to discuss the role of AI in sustainability. We explore real-world use cases where AI-driven solutions are leveraged to help tackle environmental and societal challenges, from early warning systems for extreme weather events to reducing food waste along the supply chain to conserving the Amazon rainforest. We cover the major threats that sustainability aims to address, the complexities in standardized sustainability compliance reporting, and the factors driving businesses to take a step toward sustainable practices. Lastly, Andres addresses the ways LLMs and generative AI can be applied towards the challenges of sustainability.

The complete show notes for this episode can be found at https://twimlai.com/go/689.

Today we’re joined by Fatih Porikli, senior director of technology at Qualcomm AI Research. In our conversation, we covered several of the Qualcomm team’s 16 accepted main track and workshop papers at this year’s CVPR conference. The papers span a variety of generative AI and traditional computer vision topics, with an emphasis on increased training and inference efficiency for mobile and edge deployment. We explore efficient diffusion models for text-to-image generation, grounded reasoning in videos using language models, real-time on-device 360° image generation for video portrait relighting, unique video-language model for situated interactions like fitness coaching, and visual reasoning model and benchmark for interpreting complex mathematical plots, and more! We also touched on several of the demos the team will be presenting at the conference, including multi-modal vision-language models (LLaVA) and parameter-efficient fine tuning (LoRA) on mobile phones.

The complete show notes for this episode can be found at https://twimlai.com/go/688.

Today, we're joined by Sasha Luccioni, AI and Climate lead at Hugging Face, to discuss the environmental impact of AI models. We dig into her recent research into the relative energy consumption of general purpose pre-trained models vs. task-specific, non-generative models for common AI tasks. We discuss the implications of the significant difference in efficiency and power consumption between the two types of models. Finally, we explore the complexities of energy efficiency and performance benchmarking, and talk through Sasha’s recent initiative, Energy Star Ratings for AI Models, a rating system designed to help AI users select and deploy models based on their energy efficiency.

The complete show notes for this episode can be found at http://twimlai.com/go/687.

Today, we're joined by Christopher Manning, the Thomas M. Siebel professor in Machine Learning at Stanford University and a recent recipient of the 2024 IEEE John von Neumann medal. In our conversation with Chris, we discuss his contributions to foundational research areas in NLP, including word embeddings and attention. We explore his perspectives on the intersection of linguistics and large language models, their ability to learn human language structures, and their potential to teach us about human language acquisition. We also dig into the concept of “intelligence” in language models, as well as the reasoning capabilities of LLMs. Finally, Chris shares his current research interests, alternative architectures he anticipates emerging beyond the LLM, and opportunities ahead in AI research.

The complete show notes for this episode can be found at https://twimlai.com/go/686.

Today we're joined by Abdul Fatir Ansari, a machine learning scientist at AWS AI Labs in Berlin, to discuss his paper, "Chronos: Learning the Language of Time Series." Fatir explains the challenges of leveraging pre-trained language models for time series forecasting. We explore the advantages of Chronos over statistical models, as well as its promising results in zero-shot forecasting benchmarks. Finally, we address critiques of Chronos, the ongoing research to improve synthetic data quality, and the potential for integrating Chronos into production systems.

The complete show notes for this episode can be found at twimlai.com/go/685.

Today we're joined by Joel Hestness, principal research scientist and lead of the core machine learning team at Cerebras. We discuss Cerebras’ custom silicon for machine learning, Wafer Scale Engine 3, and how the latest version of the company’s single-chip platform for ML has evolved to support large language models. Joel shares how WSE3 differs from other AI hardware solutions, such as GPUs, TPUs, and AWS’ Inferentia, and talks through the homogenous design of the WSE chip and its memory architecture. We discuss software support for the platform, including support by open source ML frameworks like Pytorch, and support for different types of transformer-based models. Finally, Joel shares some of the research his team is pursuing to take advantage of the hardware's unique characteristics, including weight-sparse training, optimizers that leverage higher-order statistics, and more.

The complete show notes for this episode can be found at twimlai.com/go/684.

Today we're joined by Laurent Boinot, power and utilities lead for the Americas at Microsoft, to discuss the intersection of AI and energy infrastructure. We discuss the many challenges faced by current power systems in North America and the role AI is beginning to play in driving efficiencies in areas like demand forecasting and grid optimization. Laurent shares a variety of examples along the way, including some of the ways utility companies are using AI to ensure secure systems, interact with customers, navigate internal knowledge bases, and design electrical transmission systems. We also discuss the future of nuclear power, and why electric vehicles might play a critical role in American energy management.

The complete show notes for this episode can be found at twimlai.com/go/683.

Today we're joined by Azarakhsh (Aza) Jalalvand, a research scholar at Princeton University, to discuss his work using deep reinforcement learning to control plasma instabilities in nuclear fusion reactors. Aza explains his team developed a model to detect and avoid a fatal plasma instability called ‘tearing mode’. Aza walks us through the process of collecting and pre-processing the complex diagnostic data from fusion experiments, training the models, and deploying the controller algorithm on the DIII-D fusion research reactor. He shares insights from developing the controller and discusses the future challenges and opportunities for AI in enabling stable and efficient fusion energy production.

The complete show notes for this episode can be found at twimlai.com/go/682.

Today we're joined by Kirk Marple, CEO and founder of Graphlit, to explore the emerging paradigm of "GraphRAG," or Graph Retrieval Augmented Generation. In our conversation, Kirk digs into the GraphRAG architecture and how Graphlit uses it to offer a multi-stage workflow for ingesting, processing, retrieving, and generating content using LLMs (like GPT-4) and other Generative AI tech. He shares how the system performs entity extraction to build a knowledge graph and how graph, vector, and object storage are integrated in the system. We dive into how the system uses “prompt compilation” to improve the results it gets from Large Language Models during generation. We conclude by discussing several use cases the approach supports, as well as future agent-based applications it enables.

The complete show notes for this episode can be found at twimlai.com/go/681.

Today we're joined by Alex Havrilla, a PhD student at Georgia Tech, to discuss "Teaching Large Language Models to Reason with Reinforcement Learning." Alex discusses the role of creativity and exploration in problem solving and explores the opportunities presented by applying reinforcement learning algorithms to the challenge of improving reasoning in large language models. Alex also shares his research on the effect of noise on language model training, highlighting the robustness of LLM architecture. Finally, we delve into the future of RL, and the potential of combining language models with traditional methods to achieve more robust AI reasoning.

The complete show notes for this episode can be found at twimlai.com/go/680.

Today we're joined by Peter Hase, a fifth-year PhD student at the University of North Carolina NLP lab. We discuss "scalable oversight", and the importance of developing a deeper understanding of how large neural networks make decisions. We learn how matrices are probed by interpretability researchers, and explore the two schools of thought regarding how LLMs store knowledge. Finally, we discuss the importance of deleting sensitive information from model weights, and how "easy-to-hard generalization" could increase the risk of releasing open-source foundation models.

The complete show notes for this episode can be found at twimlai.com/go/679.

Today we're joined by Jonas Geiping, a research group leader at the ELLIS Institute, to explore his paper: "Coercing LLMs to Do and Reveal (Almost) Anything". Jonas explains how neural networks can be exploited, highlighting the risk of deploying LLM agents that interact with the real world. We discuss the role of open models in enabling security research, the challenges of optimizing over certain constraints, and the ongoing difficulties in achieving robustness in neural networks. Finally, we delve into the future of AI security, and the need for a better approach to mitigate the risks posed by optimized adversarial attacks.

The complete show notes for this episode can be found at twimlai.com/go/678.

Today we’re joined by Mido Assran, a research scientist at Meta’s Fundamental AI Research (FAIR). In this conversation, we discuss V-JEPA, a new model being billed as “the next step in Yann LeCun's vision” for true artificial reasoning. V-JEPA, the video version of Meta’s Joint Embedding Predictive Architecture, aims to bridge the gap between human and machine intelligence by training models to learn abstract concepts in a more efficient predictive manner than generative models. V-JEPA uses a novel self-supervised training approach that allows it to learn from unlabeled video data without being distracted by pixel-level detail. Mido walks us through the process of developing the architecture and explains why it has the potential to revolutionize AI.

The complete show notes for this episode can be found at twimlai.com/go/677.

Today we’re joined by Sherry Yang, senior research scientist at Google DeepMind and a PhD student at UC Berkeley. In this interview, we discuss her new paper, "Video as the New Language for Real-World Decision Making,” which explores how generative video models can play a role similar to language models as a way to solve tasks in the real world. Sherry draws the analogy between natural language as a unified representation of information and text prediction as a common task interface and demonstrates how video as a medium and generative video as a task exhibit similar properties. This formulation enables video generation models to play a variety of real-world roles as planners, agents, compute engines, and environment simulators. Finally, we explore UniSim, an interactive demo of Sherry's work and a preview of her vision for interacting with AI-generated environments.

The complete show notes for this episode can be found at twimlai.com/go/676.

Today we’re joined by Sayash Kapoor, a Ph.D. student in the Department of Computer Science at Princeton University. Sayash walks us through his paper: "On the Societal Impact of Open Foundation Models.” We dig into the controversy around AI safety, the risks and benefits of releasing open model weights, and how we can establish common ground for assessing the threats posed by AI. We discuss the application of the framework presented in the paper to specific risks, such as the biosecurity risk of open LLMs, as well as the growing problem of "Non Consensual Intimate Imagery" using open diffusion models.

The complete show notes for this episode can be found at twimlai.com/go/675.

Today we’re joined by Akshita Bhagia, a senior research engineer at the Allen Institute for AI. Akshita joins us to discuss OLMo, a new open source language model with 7 billion and 1 billion variants, but with a key difference compared to similar models offered by Meta, Mistral, and others. Namely, the fact that AI2 has also published the dataset and key tools used to train the model. In our chat with Akshita, we dig into the OLMo models and the various projects falling under the OLMo umbrella, including Dolma, an open three-trillion-token corpus for language model pretraining, and Paloma, a benchmark and tooling for evaluating language model performance across a variety of domains.

The complete show notes for this episode can be found at twimlai.com/go/674.

Today we’re joined by Ben Prystawski, a PhD student in the Department of Psychology at Stanford University working at the intersection of cognitive science and machine learning. Our conversation centers on Ben’s recent paper, “Why think step by step? Reasoning emerges from the locality of experience,” which he recently presented at NeurIPS 2023. In this conversation, we start out exploring basic questions about LLM reasoning, including whether it exists, how we can define it, and how techniques like chain-of-thought reasoning appear to strengthen it. We then dig into the details of Ben’s paper, which aims to understand why thinking step-by-step is effective and demonstrates that local structure is the key property of LLM training data that enables it.

The complete show notes for this episode can be found at twimlai.com/go/673.

Today we're joined by Armineh Nourbakhsh of JP Morgan AI Research to discuss the development and capabilities of DocLLM, a layout-aware large language model for multimodal document understanding. Armineh provides a historical overview of the challenges of document AI and an introduction to the DocLLM model. Armineh explains how this model, distinct from both traditional LLMs and document AI models, incorporates both textual semantics and spatial layout in processing enterprise documents like reports and complex contracts. We dig into her team’s approach to training DocLLM, their choice of a generative model as opposed to an encoder-based approach, the datasets they used to build the model, their approach to incorporating layout information, and the various ways they evaluated the model’s performance.

The complete show notes for this episode can be found at twimlai.com/go/672.

Today we’re joined by Sanmi Koyejo, assistant professor at Stanford University, to continue our NeurIPS 2024 series. In our conversation, Sanmi discusses his two recent award-winning papers. First, we dive into his paper, “Are Emergent Abilities of Large Language Models a Mirage?”. We discuss the different ways LLMs are evaluated and the excitement surrounding their“emergent abilities” such as the ability to perform arithmetic Sanmi describes how evaluating model performance using nonlinear metrics can lead to the illusion that the model is rapidly gaining new capabilities, whereas linear metrics show smooth improvement as expected, casting doubt on the significance of emergence. We continue on to his next paper, “DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT Models,” discussing the methodology it describes for evaluating concerns such as the toxicity, privacy, fairness, and robustness of LLMs.

The complete show notes for this episode can be found at twimlai.com/go/671.

Today we’re joined by Kamyar Azizzadenesheli, a staff researcher at Nvidia, to continue our AI Trends 2024 series. In our conversation, Kamyar updates us on the latest developments in reinforcement learning (RL), and how the RL community is taking advantage of the abstract reasoning abilities of large language models (LLMs). Kamyar shares his insights on how LLMs are pushing RL performance forward in a variety of applications, such as ALOHA, a robot that can learn to fold clothes, and Voyager, an RL agent that uses GPT-4 to outperform prior systems at playing Minecraft. We also explore the progress being made in assessing and addressing the risks of RL-based decision-making in domains such as finance, healthcare, and agriculture. Finally, we discuss the future of deep reinforcement learning, Kamyar’s top predictions for the field, and how greater compute capabilities will be critical in achieving general intelligence.

The complete show notes for this episode can be found at twimlai.com/go/670.

Today we’re joined by Ram Sriharsha, VP of engineering at Pinecone. In our conversation, we dive into the topic of vector databases and retrieval augmented generation (RAG). We explore the trade-offs between relying solely on LLMs for retrieval tasks versus combining retrieval in vector databases and LLMs, the advantages and complexities of RAG with vector databases, the key considerations for building and deploying real-world RAG-based applications, and an in-depth look at Pinecone's new serverless offering. Currently in public preview, Pinecone Serverless is a vector database that enables on-demand data loading, flexible scaling, and cost-effective query processing. Ram discusses how the serverless paradigm impacts the vector database’s core architecture, key features, and other considerations. Lastly, Ram shares his perspective on the future of vector databases in helping enterprises deliver RAG systems.

The complete show notes for this episode can be found at twimlai.com/go/669.

Today we’re joined by Ben Zhao, a Neubauer professor of computer science at the University of Chicago. In our conversation, we explore his research at the intersection of security and generative AI. We focus on Ben’s recent Fawkes, Glaze, and Nightshade projects, which use “poisoning” approaches to provide users with security and protection against AI encroachments. The first tool we discuss, Fawkes, imperceptibly “cloaks” images in such a way that models perceive them as highly distorted, effectively shielding individuals from recognition by facial recognition models. We then dig into Glaze, a tool that employs machine learning algorithms to compute subtle alterations that are indiscernible to human eyes but adept at tricking the models into perceiving a significant shift in art style, giving artists a unique defense against style mimicry. Lastly, we cover Nightshade, a strategic defense tool for artists akin to a 'poison pill' which allows artists to apply imperceptible changes to their images that effectively “breaks” generative AI models that are trained on them.

The complete show notes for this episode can be found at twimlai.com/go/668.

Today, we continue our NeurIPS series with Dan Friedman, a PhD student in the Princeton NLP group. In our conversation, we explore his research on mechanistic interpretability for transformer models, specifically his paper, Learning Transformer Programs. The LTP paper proposes modifications to the transformer architecture which allow transformer models to be easily converted into human-readable programs, making them inherently interpretable. In our conversation, we compare the approach proposed by this research with prior approaches to understanding the models and their shortcomings. We also dig into the approach’s function and scale limitations and constraints.

The complete show notes for this episode can be found at twimlai.com/go/667.

Today we continue our AI Trends 2024 series with a conversation with Thomas Dietterich, distinguished professor emeritus at Oregon State University. As you might expect, Large Language Models figured prominently in our conversation, and we covered a vast array of papers and use cases exploring current research into topics such as monolithic vs. modular architectures, hallucinations, the application of uncertainty quantification (UQ), and using RAG as a sort of memory module for LLMs. Lastly, don’t miss Tom’s predictions on what he foresees happening this year as well as his words of encouragement for those new to the field.

The complete show notes for this episode can be found at twimlai.com/go/666.

Today we kick off our AI Trends 2024 series with a conversation with Naila Murray, director of AI research at Meta. In our conversation with Naila, we dig into the latest trends and developments in the realm of computer vision. We explore advancements in the areas of controllable generation, visual programming, 3D Gaussian splatting, and multimodal models, specifically vision plus LLMs. We discuss tools and open source projects, including Segment Anything–a tool for versatile zero-shot image segmentation using simple text prompts clicks, and bounding boxes; ControlNet–which adds conditional control to stable diffusion models; and DINOv2–a visual encoding model enabling object recognition, segmentation, and depth estimation, even in data-scarce scenarios. Finally, Naila shares her view on the most exciting opportunities in the field, as well as her predictions for upcoming years.

The complete show notes for this episode can be found at twimlai.com/go/665.

Today we’re joined by Ed Anuff, chief product officer at DataStax. In our conversation, we discuss Ed’s insights on RAG, vector databases, embedding models, and more. We dig into the underpinnings of modern vector databases (like HNSW and DiskANN) that allow them to efficiently handle massive and unstructured data sets, and discuss how they help users serve up relevant results for RAG, AI assistants, and other use cases. We also discuss embedding models and their role in vector comparisons and database retrieval as well as the potential for GPU usage to enhance vector database performance.

The complete show notes for this episode can be found at twimlai.com/go/664.

Today we’re joined by Markus Nagel, research scientist at Qualcomm AI Research, who helps us kick off our coverage of NeurIPS 2023. In our conversation with Markus, we cover his accepted papers at the conference, along with other work presented by Qualcomm AI Research scientists. Markus’ first paper, Quantizable Transformers: Removing Outliers by Helping Attention Heads Do Nothing, focuses on tackling activation quantization issues introduced by the attention mechanism and how to solve them. We also discuss Pruning vs Quantization: Which is Better?, which focuses on comparing the effectiveness of these two methods in achieving model weight compression. Additional papers discussed focus on topics like using scalarization in multitask and multidomain learning to improve training and inference, using diffusion models for a sequence of state models and actions, applying geometric algebra with equivariance to transformers, and applying a deductive verification of chain of thought reasoning performed by LLMs.

The complete show notes for this episode can be found at twimlai.com/go/663.

Today we’re joined by Michael Kearns, professor in the Department of Computer and Information Science at the University of Pennsylvania and an Amazon scholar. In our conversation with Michael, we discuss the new challenges to responsible AI brought about by the generative AI era. We explore Michael’s learnings and insights from the intersection of his real-world experience at AWS and his work in academia. We cover a diverse range of topics under this banner, including service card metrics, privacy, hallucinations, RLHF, and LLM evaluation benchmarks. We also touch on Clean Rooms ML, a secured environment that balances accessibility to private datasets through differential privacy techniques, offering a new approach for secure data handling in machine learning.

The complete show notes for this episode can be found at twimlai.com/go/662.

Today we’re joined by Mike Miller, director of product at AWS responsible for the company’s “edutainment” products. In our conversation with Mike, we explore AWS PartyRock, a no-code generative AI app builder that allows users to easily create fun and shareable AI applications by selecting a model, chaining prompts together, and linking different text, image, and chatbot widgets together. Additionally, we discuss some of the previous tools Mike’s team has delivered at the intersection of developer education and entertainment, including DeepLens, a computer vision hardware device, DeepRacer, a programmable vehicle that uses reinforcement learning to navigate a track, and lastly, DeepComposer, a generative AI model that transforms musical inputs and creates accompanying compositions.

The complete show notes for this episode can be found at twimlai.com/go/661.

Today we’re joined by Cody Coleman, co-founder and CEO of Coactive AI. In our conversation with Cody, we discuss how Coactive has leveraged modern data, systems, and machine learning techniques to deliver its multimodal asset platform and visual search tools. Cody shares his expertise in the area of data-centric AI, and we dig into techniques like active learning and core set selection, and how they can drive greater efficiency throughout the machine learning lifecycle. We explore the various ways Coactive uses multimodal embeddings to enable their core visual search experience, and we cover the infrastructure optimizations they’ve implemented in order to scale their systems. We conclude with Cody’s advice for entrepreneurs and engineers building companies around generative AI technologies.

The complete show notes for this episode can be found at twimlai.com/go/660.

Today we’re joined by Kyle Roche, founder and CEO of Griptape to discuss patterns and middleware for LLM applications. We dive into the emerging patterns for developing LLM applications, such as off prompt data—which allows data retrieval without compromising the chain of thought within language models—and pipelines, which are sequential tasks that are given to LLMs that can involve different models for each task or step in the pipeline. We also explore Griptape, an open-source, Python-based middleware stack that aims to securely connect LLM applications to an organization’s internal and external data systems. We discuss the abstractions it offers, including drivers, memory management, rule sets, DAG-based workflows, and a prompt stack. Additionally, we touch on common customer concerns such as privacy, retraining, and sovereignty issues, and several use cases that leverage role-based retrieval methods to optimize human augmentation tasks.

The complete show notes for this episode can be found at twimlai.com/go/659.

Today we’re joined by Prem Natarajan, chief scientist and head of enterprise AI at Capital One. In our conversation, we discuss AI access and inclusivity as technical challenges and explore some of Prem and his team’s multidisciplinary approaches to tackling these complexities. We dive into the issues of bias, dealing with class imbalances, and the integration of various research initiatives to achieve additive results. Prem also shares his team’s work on foundation models for financial data curation, highlighting the importance of data quality and the use of federated learning, and emphasizing the impact these factors have on the model performance and reliability in critical applications like fraud detection. Lastly, Prem shares his overall approach to tackling AI research in the context of a banking enterprise, including prioritizing mission-inspired research aiming to deliver tangible benefits to customers and the broader community, investing in diverse talent and the best infrastructure, and forging strategic partnerships with a variety of academic labs.

The complete show notes for this episode can be found at twimlai.com/go/658.

Today we’re joined by Jay Emery, director of technical sales & architecture at Microsoft Azure. In our conversation with Jay, we discuss the challenges faced by organizations when building LLM-based applications, and we explore some of the techniques they are using to overcome them. We dive into the concerns around security, data privacy, cost management, and performance as well as the ability and effectiveness of prompting to achieve the desired results versus fine-tuning, and when each approach should be applied. We cover methods such as prompt tuning and prompt chaining, prompt variance, fine-tuning, and RAG to enhance LLM output along with ways to speed up inference performance such as choosing the right model, parallelization, and provisioned throughput units (PTUs). In addition to that, Jay also shared several intriguing use cases describing how businesses use tools like Azure Machine Learning prompt flow and Azure ML AI Studio to tailor LLMs to their unique needs and processes.

The complete show notes for this episode can be found at twimlai.com/go/657.

Today we’re joined by Richard Zhang, senior research scientist at Adobe Research. In our conversation with Richard, we explore the research challenges that arise when regarding visual generative AI from an ecosystem perspective, considering the disparate needs of creators, consumers, and contributors. We start with his work on perceptual metrics and the LPIPS paper, which allow us to better align human perception and computer vision and which remain used in contemporary generative AI applications such as stable diffusion, GANs, and latent diffusion. We look at his work creating detection tools for fake visual content, highlighting the importance of generalization of these detection methods to new, unseen models. Lastly, we dig into his work on data attribution and concept ablation, which aim to address the challenging open problem of allowing artists and others to manage their contributions to generative AI training data sets.

The complete show notes for this episode can be found at twimlai.com/go/656.

Today we’re joined by Heather Gorr, principal MATLAB product marketing manager at MathWorks. In our conversation with Heather, we discuss the deployment of AI models to hardware devices and embedded AI systems. We explore factors to consider during data preparation, model development, and ultimately deployment, to ensure a successful project. Factors such as device constraints and latency requirements which dictate the amount and frequency of data flowing onto the device are discussed, as are modeling needs such as explainability, robustness and quantization; the use of simulation throughout the modeling process; the need to apply robust verification and validation methodologies to ensure safety and reliability; and the need to adapt and apply MLOps techniques for speed and consistency. Heather also shares noteworthy anecdotes about embedded AI deployments in industries including automotive and oil & gas.

The complete show notes for this episode can be found at twimlai.com/go/655.

Today we’re joined by Yoshua Bengio, professor at Université de Montréal. In our conversation with Yoshua, we discuss AI safety and the potentially catastrophic risks of its misuse. Yoshua highlights various risks and the dangers of AI being used to manipulate people, spread disinformation, cause harm, and further concentrate power in society. We dive deep into the risks associated with achieving human-level competence in enough areas with AI, and tackle the challenges of defining and understanding concepts like agency and sentience. Additionally, our conversation touches on solutions to AI safety, such as the need for robust safety guardrails, investments in national security protections and countermeasures, bans on systems with uncertain safety, and the development of governance-driven AI systems.

The complete show notes for this episode can be found at twimlai.com/go/654.

Today we’re joined by Miriam Friedel, senior director of ML engineering at Capital One. In our conversation with Miriam, we discuss some of the challenges faced when delivering machine learning tools and systems in highly regulated enterprise environments, and some of the practices her teams have adopted to help them operate with greater speed and agility. We also explore how to create a culture of collaboration, the value of standardized tooling and processes, leveraging open-source, and incentivizing model reuse. Miriam also shares her thoughts on building a ‘unicorn’ team, and what this means for the team she’s built at Capital One, as well as her take on build vs. buy decisions for MLOps, and the future of MLOps and enterprise AI more broadly. Throughout, Miriam shares examples of these ideas at work in some of the tools their team has built, such as Rubicon, an open source experiment management tool, and Kubeflow pipeline components that enable Capital One data scientists to efficiently leverage and scale models. 

The complete show notes for this episode can be found at twimlai.com/go/653.

Today we’re joined by Riley Goodside, staff prompt engineer at Scale AI. In our conversation with Riley, we explore LLM capabilities and limitations, prompt engineering, and the mental models required to apply advanced prompting techniques. We dive deep into understanding LLM behavior, discussing the mechanism of autoregressive inference, comparing k-shot and zero-shot prompting, and dissecting the impact of RLHF. We also discuss the idea that prompting is a scaffolding structure that leverages the model context, resulting in achieving the desired model behavior and response rather than focusing solely on writing ability.

The complete show notes for this episode can be found at twimlai.com/go/652.

Today we’re joined by Sara Hooker, director at Cohere and head of Cohere For AI, Cohere’s research lab. In our conversation with Sara, we explore some of the challenges with multilingual models like poor data quality and tokenization, and how they rely on data augmentation and preference training to address these bottlenecks. We also discuss the disadvantages and the motivating factors behind the Mixture of Experts technique, and the importance of common language between ML researchers and hardware architects to address the pain points in frameworks and create a better cohesion between the distinct communities. Sara also highlights the impact and the emotional connection that language models have created in society, the benefits and the current safety concerns of universal models, and the significance of having grounded conversations to characterize and mitigate the risk and development of AI models. Along the way, we also dive deep into Cohere and Cohere for AI, along with their Aya project, an open science project that aims to build a state-of-the-art multilingual generative language model as well as some of their recent research papers.

The complete show notes for this episode can be found at twimlai.com/go/651.

Today we’re joined by Luke Zettlemoyer, professor at University of Washington and a research manager at Meta. In our conversation with Luke, we cover multimodal generative AI, the effect of data on models, and the significance of open source and open science. We explore the grounding problem, the need for visual grounding and embodiment in text-based models, the advantages of discretization tokenization in image generation, and his paper Scaling Laws for Generative Mixed-Modal Language Models, which focuses on simultaneously training LLMs on various modalities. Additionally, we cover his papers on Self-Alignment with Instruction Backtranslation, and LIMA: Less Is More for Alignment.

The complete show notes for this episode can be found at twimlai.com/go/650.

Today we’re joined by Alex Hanna, the Director of Research at the Distributed AI Research Institute (DAIR). In our conversation with Alex, we discuss the topic of AI hype and the importance of tackling the issues and impacts it has on society. Alex highlights how the hype cycle started, concerning use cases, incentives driving people towards the rapid commercialization of AI tools, and the need for robust evaluation tools and frameworks to assess and mitigate the risks of these technologies. We also talked about DAIR and how they’ve crafted their research agenda. We discuss current research projects like DAIR Fellow Asmelash Teka Hadgu’s research supporting machine translation and speech recognition tools for the low-resource Amharic and Tigrinya languages of Ethiopia and Eritrea, in partnership with his startup Lesan.AI. We also explore the “Do Data Sets Have Politics” paper, which focuses on coding various variables and conducting a qualitative analysis of computer vision data sets to uncover the inherent politics present in data sets and the challenges in data set creation.

The complete show notes for this episode can be found at twimlai.com/go/649.

Today we’re joined by Nataniel Ruiz, a research scientist at Google. In our conversation with Nataniel, we discuss his recent work around personalization for text-to-image AI models. Specifically, we dig into DreamBooth, an algorithm that enables “subject-driven generation,” that is, the creation of personalized generative models using a small set of user-provided images about a subject. The personalized models can then be used to generate the subject in various contexts using a text prompt. Nataniel gives us a dive deep into the fine-tuning approach used in DreamBooth, the potential reasons behind the algorithm’s effectiveness, the challenges of fine-tuning diffusion models in this way, such as language drift, and how the prior preservation loss technique avoids this setback, as well as the evaluation challenges and metrics used in DreamBooth. We also touched base on his other recent papers including SuTI, StyleDrop, HyperDreamBooth, and lastly, Platypus.

The complete show notes for this episode can be found at twimlai.com/go/648.

Today we’re joined by Shreya Rajpal, founder and CEO of Guardrails AI. In our conversation with Shreya, we discuss ensuring the safety and reliability of language models for production applications. We explore the risks and challenges associated with these models, including different types of hallucinations and other LLM failure modes. We also talk about the susceptibility of the popular retrieval augmented generation (RAG) technique to closed-domain hallucination, and how this challenge can be addressed. We also cover the need for robust evaluation metrics and tooling for building with large language models. Lastly, we explore Guardrails, an open-source project that provides a catalog of validators that run on top of language models to enforce correctness and reliability efficiently.

The complete show notes for this episode can be found at twimlai.com/go/647.

Today we’re joined by Roland Memisevic, a senior director at Qualcomm AI Research. In our conversation with Roland, we discuss the significance of language in humanlike AI systems and the advantages and limitations of autoregressive models like Transformers in building them. We cover the current and future role of recurrence in LLM reasoning and the significance of improving grounding in AI—including the potential of developing a sense of self in agents. Along the way, we discuss Fitness Ally, a fitness coach trained on a visually grounded large language model, which has served as a platform for Roland’s research into neural reasoning, as well as recent research that explores topics like visual grounding for large language models and state-augmented architectures for AI agents.

The complete show notes for this episode can be found at twimlai.com/go/646.

Today we’re joined by James Zou, an assistant professor at Stanford University. In our conversation with James, we explore the differences in ChatGPT’s behavior over the last few months. We discuss the issues that can arise from inconsistencies in generative AI models, how he tested ChatGPT’s performance in various tasks, drawing comparisons between March 2023 and June 2023 for both GPT-3.5 and GPT-4 versions, and the possible reasons behind the declining performance of these models. James also shared his thoughts on how surgical AI editing akin to CRISPR could potentially revolutionize LLM and AI systems, and how adding monitoring tools can help in tracking behavioral changes in these models. Finally, we discuss James' recent paper on pathology image analysis using Twitter data, in which he explores the challenges of obtaining large medical datasets and data collection, as well as detailing the model’s architecture, training, and the evaluation process.

The complete show notes for this episode can be found at twimlai.com/go/645.

Today we’re joined by Sophia Sanborn, a postdoctoral scholar at the University of California, Santa Barbara. In our conversation with Sophia, we explore the concept of universality between neural representations and deep neural networks, and how these principles of efficiency provide an ability to find consistent features across networks and tasks. We also discuss her recent paper on Bispectral Neural Networks which focuses on Fourier transform and its relation to group theory, the implementation of bi-spectral spectrum in achieving invariance in deep neural networks, the expansion of geometric deep learning on the concept of CNNs from other domains, the similarities in the fundamental structure of artificial neural networks and biological neural networks and how applying similar constraints leads to the convergence of their solutions.

The complete show notes for this episode can be found at twimlai.com/go/644.

Today we’re joined by Gokul Swamy, a Ph.D. Student at the Robotics Institute at Carnegie Mellon University. In the final conversation of our ICML 2023 series, we sat down with Gokul to discuss his accepted papers at the event, leading off with “Inverse Reinforcement Learning without Reinforcement Learning.” In this paper, Gokul explores the challenges and benefits of inverse reinforcement learning, and the potential and advantages it holds for various applications. Next up, we explore the “Complementing a Policy with a Different Observation Space” paper which applies causal inference techniques to accurately estimate sampling balance and make decisions based on limited observed features. Finally, we touched on “Learning Shared Safety Constraints from Multi-task Demonstrations” which centers on learning safety constraints from demonstrations using the inverse reinforcement learning approach.

The complete show notes for this episode can be found at twimlai.com/go/643.

Today we’re joined by Su-In Lee, a professor at the Paul G. Allen School of Computer Science And Engineering at the University Of Washington. In our conversation, Su-In details her talk from the ICML 2023 Workshop on Computational Biology which focuses on developing explainable AI techniques for the computational biology and clinical medicine fields. Su-In discussed the importance of explainable AI contributing to feature collaboration, the robustness of different explainability approaches, and the need for interdisciplinary collaboration between the computer science, biology, and medical fields. We also explore her recent paper on the use of drug combination therapy, challenges with handling biomedical data, and how they aim to make meaningful contributions to the healthcare industry by aiding in cause identification and treatments for Cancer and Alzheimer's diseases.

The complete show notes for this episode can be found at twimlai.com/go/642.

Today we’re joined by Bayan Bruss, Vice President of Applied ML Research at Capital One. In our conversation with Bayan, we covered a pair of papers his team presented at this year’s ICML conference. We begin with the paper Interpretable Subspaces in Image Representations, where Bayan gives us a dive deep into the interpretability framework, embedding dimensions, contrastive approaches, and how their model can accelerate image representation in deep learning. We also explore GOAT: A Global Transformer on Large-scale Graphs, a scalable global graph transformer. We talk through the computation challenges, homophilic and heterophilic principles, model sparsity, and how their research proposes methodologies to get around the computational barrier when scaling to large-scale graph models.

The complete show notes for this episode can be found at twimlai.com/go/641.

Today we’re joined by Atul Deo, General Manager of Amazon Bedrock. In our conversation with Atul, we discuss the process of training large language models in the enterprise, including the pain points of creating and training machine learning models, and the power of pre-trained models. We explore different approaches to how companies can leverage large language models, dealing with the hallucination, and the transformative process of retrieval augmented generation (RAG). Finally, Atul gives us an inside look at Bedrock, a fully managed service that simplifies the deployment of generative AI-based apps at scale.

The complete show notes for this episode can be found at twimlai.com/go/640.

Today we’re joined by David Rosenberg, head of the machine learning strategy team in the Office of the CTO at Bloomberg. In our conversation with David, we discuss the creation of BloombergGPT, a custom-built LLM focused on financial applications. We explore the model’s architecture, validation process, benchmarks, and its distinction from other language models. David also discussed the evaluation process, performance comparisons, progress, and the future directions of the model. Finally, we discuss the ethical considerations that come with building these types of models, and how they've approached dealing with these issues.

The complete show notes for this episode can be found at twimlai.com/go/639

Today we’re joined by Robert Osazuwa Ness, a senior researcher at Microsoft Research, Professor at Northeastern University, and Founder of Altdeep.ai. In our conversation with Robert, we explore whether large language models, specifically GPT-3, 3.5, and 4, are good at causal reasoning. We discuss the benchmarks used to evaluate these models and the limitations they have in answering specific causal reasoning questions, while Robert highlights the need for access to weights, training data, and architecture to correctly answer these questions. The episode discusses the challenge of generalization in causal relationships and the importance of incorporating inductive biases, explores the model's ability to generalize beyond the provided benchmarks, and the importance of considering causal factors in decision-making processes.

The complete show notes for this episode can be found at twimlai.com/go/638.

Today we’re joined by Alice Xiang, Lead Research Scientist at Sony AI, and Global Head of AI Ethics at Sony Group Corporation. In our conversation with Alice, we discuss the ongoing debate between privacy and fairness in computer vision, diving into the impact of data privacy laws on the AI space while highlighting concerns about unauthorized use and lack of transparency in data usage. We explore the potential harm of inaccurate AI model outputs and the need for legal protection against biased AI products, and Alice suggests various solutions to address these challenges, such as working through third parties for data collection and establishing closer relationships with communities. Finally, we talk through the history of unethical data collection practices in CV and the emergence of generative AI technologies that exacerbate the problem, the importance of operationalizing ethical data collection and practice, including appropriate consent, representation, diversity, and compensation, and the need for interdisciplinary collaboration in AI ethics and the growing interest in AI regulation, including the EU AI Act and regulatory activities in the US.

The complete show notes for this episode can be found at twimlai.com/go/637.

Today we're joined by Mohit Bansal, Parker Professor, and Director of the MURGe-Lab at UNC, Chapel Hill. In our conversation with Mohit, we explore the concept of unification in AI models, highlighting the advantages of shared knowledge and efficiency. He addresses the challenges of evaluation in generative AI, including biases and spurious correlations. Mohit introduces groundbreaking models such as UDOP and VL-T5, which achieved state-of-the-art results in various vision and language tasks while using fewer parameters. Finally, we discuss the importance of data efficiency, evaluating bias in models, and the future of multimodal models and explainability.

The complete show notes for this episode can be found at twimlai.com/go/636.

Today we kick off our coverage of the 2023 CVPR conference joined by Fatih Porikli, a Senior Director of Technology at Qualcomm. In our conversation with Fatih, we covered quite a bit of ground, touching on a total of 12 papers/demos, focusing on topics like data augmentation and optimized architectures for computer vision. We explore advances in optical flow estimation networks, cross-model, and stage knowledge distillation for efficient 3D object detection, and zero-shot learning via language models for fine-grained labeling. We also discuss generative AI advancements and computer vision optimization for running large models on edge devices. Finally, we discuss objective functions, architecture design choices for neural networks, and efficiency and accuracy improvements in AI models via the techniques introduced in the papers.

Today we’re joined by Chris Lattner, Co-Founder and CEO of Modular. In our conversation with Chris, we discuss Mojo, a new programming language for AI developers. Mojo is unique in this space and simplifies things by making the entire stack accessible and understandable to people who are not compiler engineers. It also offers Python programmers the ability to make it high-performance and capable of running accelerators, making it more accessible to more people and researchers. We discuss the relationship between the Modular Engine and Mojo, the challenge of packaging Python, particularly when incorporating C code, and how Mojo aims to solve these problems to make the AI stack more dependable.

The complete show notes for this episode can be found at twimlai.com/go/634

Today we’re joined by Jilei Hou, a VP of Engineering at Qualcomm Technologies. In our conversation with Jilei, we focus on the emergence of generative AI, and how they've worked towards providing these models for use on edge devices. We explore how the distribution of models on devices can help amortize large models' costs while improving reliability and performance and the challenges of running machine learning workloads on devices, including model size and inference latency. Finally, Jilei we explore how these emerging technologies fit into the existing AI Model Efficiency Toolkit (AIMET) framework. 

The complete show notes for this episode can be found at twimlai.com/go/633

Today we’re joined by Joon Sung Park, a PhD Student at Stanford University. Joon shares his passion for creating AI systems that can solve human problems and his work on the recent paper Generative Agents: Interactive Simulacra of Human Behavior, which showcases generative agents that exhibit believable human behavior. We discuss using empirical methods to study these systems and the conflicting papers on whether AI models have a worldview and common sense. Joon talks about the importance of context and environment in creating believable agent behavior and shares his team's work on scaling emerging community behaviors. He also dives into the importance of a long-term memory module in agents and the use of knowledge graphs in retrieving associative information. The goal, Joon explains, is to create something that people can enjoy and empower people, solving existing problems and challenges in the traditional HCI and AI field.

Today we’re joined by Hugo Larochelle, a research scientist at Google Deepmind. In our conversation with Hugo, we discuss his work on transfer learning, understanding the capabilities of deep learning models, and creating the Transactions on Machine Learning Research journal. We explore the use of large language models in NLP, prompting, and zero-shot learning. Hugo also shares insights from his research on neural knowledge mobilization for code completion and discusses the adaptive prompts used in their system. 

The complete show notes for this episode can be found at twimlai.com/go/631.

Today we’re joined by Dan Fu, a PhD student at Stanford University. In our conversation with Dan, we discuss the limitations of state space models in language modeling and the search for alternative building blocks that can help increase context length without being computationally infeasible. Dan walks us through the H3 architecture and Flash Attention technique, which can reduce the memory footprint of a model and make it feasible to fine-tune. We also explore his work on improving language models using synthetic languages, the issue of long sequence length affecting both training and inference in models, and the hope for finding something sub-quadratic that can perform language processing more effectively than the brute force approach of attention.

The complete show notes for this episode can be found at https://twimlai.com/go/630

Today we continue our coverage of ICLR 2023 joined by Dhruv Batra, an associate professor at Georgia Tech and research director of the Fundamental AI Research (FAIR) team at META. In our conversation, we discuss Dhruv’s work on the paper Emergence of Maps in the Memories of Blind Navigation Agents, which won an Outstanding Paper Award at the event. We explore navigation with multilayer LSTM and the question of whether embodiment is necessary for intelligence. We delve into the Embodiment Hypothesis and the progress being made in language models and caution on the responsible use of these models. We also discuss the history of AI and the importance of using the right data sets in training. The conversation explores the different meanings of "maps" across AI and cognitive science fields, Dhruv’s experience in navigating mapless systems, and the early discovery stages of memory representation and neural mechanisms.

The complete show notes for this episode can be found at https://twimlai.com/go/629

Today we’re joined by Jerry Liu, co-founder and CEO of Llama Index. In our conversation with Jerry, we explore the creation of Llama Index, a centralized interface to connect your external data with the latest large language models. We discuss the challenges of adding private data to language models and how Llama Index connects the two for better decision-making. We discuss the role of agents in automation, the evolution of the agent abstraction space, and the difficulties of optimizing queries over large amounts of complex data. We also discuss a range of topics from combining summarization and semantic search, to automating reasoning, to improving language model results by exploiting relationships between nodes in data. 

The complete show notes for this episode can be found at twimlai.com/go/628.

Today we kick off our coverage of the 2023 ICLR conference joined by Christos Louizos, an ML researcher at Qualcomm Technologies. In our conversation with Christos, we explore his paper Hyperparameter Optimization through Neural Network Partitioning and a few of his colleague's works from the conference. We discuss methods for speeding up attention mechanisms in transformers, scheduling operations for computation graphs, estimating channels in indoor environments, and adapting to distribution shifts in test time with neural network modules. We also talk through the benefits and limitations of federated learning, exploring sparse models, optimizing communication between servers and devices, and much more. 

The complete show notes for this episode can be found at https://twimlai.com/go/627.

Today we’re joined by Marti Hearst, Professor at UC Berkeley. In our conversation with Marti, we explore the intricacies of AI language models and their usefulness in improving efficiency but also their potential for spreading misinformation. Marti expresses skepticism about whether these models truly have cognition compared to the nuance of the human brain. We discuss the intersection of language and visualization and the need for specialized research to ensure safety and appropriateness for specific uses. We also delve into the latest tools and algorithms such as Copilot and Chat GPT, which enhance programming and help in identifying comparisons, respectively. Finally, we discuss Marti’s long research history in search and her breakthrough in developing a standard interaction that allows for finding items on websites and library catalogs.

The complete show notes for this episode can be found at https://twimlai.com/go/626.

Today we’re joined by Ben Goertzel, CEO of SingularityNET. In our conversation with Ben, we explore all things AGI, including the potential scenarios that could arise with the advent of AGI and his preference for a decentralized rollout comparable to the internet or Linux. Ben shares his research in bridging neural nets, symbolic logic engines, and evolutionary programming engines to develop a common mathematical framework for AI paradigms. We also discuss the limitations of Large Language Models and the potential of hybridizing LLMs with other AGI approaches. Additionally, we chat about their work using LLMs for music generation and the limitations of formalizing creativity. Finally, Ben discusses his team's work with the OpenCog Hyperon framework and Simuli to achieve AGI, and the potential implications of their research in the future.

The complete show notes for this episode can be found at https://twimlai.com/go/625

Today we’re joined by Jeff Boudier, head of product at Hugging Face 🤗. In our conversation with Jeff, we explore the current landscape of open-source machine learning tools and models, the recent shift towards consumer-focused releases, and the importance of making ML tools accessible. We also discuss the growth of the Hugging Face Hub, which currently hosts over 150k models, and how formalizing their collaboration with AWS will help drive the adoption of open-source models in the enterprise.  

The complete show notes for this episode can be found at twimlai.com/go/624

Today we’re joined by Vinesh Sukumar, a senior director and head of AI/ML product management at Qualcomm Technologies. In our conversation with Vinesh, we explore how mobile and automotive devices have different requirements for AI models and how their AI stack helps developers create complex models on both platforms. We also discuss the growing interest in text-based input and the shift towards transformers, generative content, and recommendation engines. Additionally, we explore the challenges and opportunities for ML Ops investments on the edge, including the use of synthetic data and evolving models based on user data. Finally, we delve into the latest advancements in large language models, including Prometheus-style models and GPT-4.

The complete show notes for this episode can be found at twimlai.com/go/623.

Today we’re joined by Anastasis Germanidis, Co-Founder and CTO of RunwayML. Amongst all the product and model releases over the past few months, Runway threw its hat into the ring with Gen-1, a model that can take still images or video and transform them into completely stylized videos. They followed that up just a few weeks later with the release of Gen-2, a multimodal model that can produce a video from text prompts. We had the pleasure of chatting with Anastasis about both models, exploring the challenges of generating video, the importance of alignment in model deployment, the potential use of RLHF, the deployment of models as APIs, and much more!

The complete show notes for this episode can be found at twimlai.com/go/622.

Today we’re joined by Tom Goldstein, an associate professor at the University of Maryland. Tom’s research sits at the intersection of ML and optimization and has previously been featured in the New Yorker for his work on invisibility cloaks, clothing that can evade object detection. In our conversation, we focus on his more recent research on watermarking LLM output. We explore the motivations behind adding these watermarks, how they work, and different ways a watermark could be deployed, as well as political and economic incentive structures around the adoption of watermarking and future directions for that line of work. We also discuss Tom’s research into data leakage, particularly in stable diffusion models, work that is analogous to recent guest Nicholas Carlini’s research into LLM data extraction. 

Today we’re joined by Anna Ivanova, a postdoctoral researcher at MIT Quest for Intelligence. In our conversation with Anna, we discuss her recent paper Dissociating language and thought in large language models: a cognitive perspective. In the paper, Anna reviews the capabilities of LLMs by considering their performance on two different aspects of language use: 'formal linguistic competence', which includes knowledge of rules and patterns of a given language, and 'functional linguistic competence', a host of cognitive abilities required for language understanding and use in the real world. We explore parallels between linguistic competence and AGI, the need to identify new benchmarks for these models, whether an end-to-end trained LLM can address various aspects of functional competence, and much more! 

The complete show notes for this episode can be found at twimlai.com/go/620.

Today we’re joined by Monroe Kennedy III, an assistant professor at Stanford, director of the Assistive Robotics and Manipulation Lab, and a national director of Black in Robotics. In our conversation with Monroe, we spend some time exploring the robotics landscape, getting Monroe’s thoughts on the current challenges in the field, as well as his opinion on choreographed demonstrations like the dancing Boston Robotics machines. We also dig into his work around two distinct threads, Robotic Dexterity, (what does it take to make robots capable of doing manipulation useful tasks with and for humans?) and Collaborative Robotics (how do we go beyond advanced autonomy in robots towards making effective robotic teammates capable of working with human counterparts?). Finally, we discuss DenseTact, an optical-tactile sensor capable of visualizing the deformed surface of a soft fingertip and using that image in a neural network to perform calibrated shape reconstruction and 6-axis wrench estimation.

The complete show notes for this episode can be found at twimlai.com/go/619.

Today we’re joined by Nicholas Carlini, a research scientist at Google Brain. Nicholas works at the intersection of machine learning and computer security, and his recent paper “Extracting Training Data from LLMs” has generated quite a buzz within the ML community. In our conversation, we discuss the current state of adversarial machine learning research, the dynamic of dealing with privacy issues in black box vs accessible models, what privacy attacks in vision models like diffusion models look like, and the scale of “memorization” within these models. We also explore Nicholas’ work on data poisoning, which looks to understand what happens if a bad actor can take control of a small fraction of the data that an ML model is trained on.

The complete show notes for this episode can be found at twimlai.com/go/618.

Today we’re joined by Vinodkumar Prabhakaran, a Senior Research Scientist at Google Research. In our conversation with Vinod, we discuss his two main areas of research, using ML, specifically NLP, to explore these social disparities, and how these same social disparities are captured and propagated within machine learning tools. We explore a few specific projects, the first using NLP to analyze interactions between police officers and community members, determining factors like level of respect or politeness and how they play out across a spectrum of community members. We also discuss his work on understanding how bias creeps into the pipeline of building ML models, whether it be from the data or the person building the model. Finally, for those working with human annotators, Vinod shares his thoughts on how to incorporate principles of fairness to help build more robust models. 

The complete show notes for this episode can be found at https://twimlai.com/go/617.

Today we’re joined by Robert Osazuwa Ness, a senior researcher at Microsoft Research, to break down the latest trends in the world of causal modeling. In our conversation with Robert, we explore advances in areas like causal discovery, causal representation learning, and causal judgements. We also discuss the impact causality could have on large language models, especially in some of the recent use cases we’ve seen like Bing Search and ChatGPT. Finally, we discuss the benchmarks for causal modeling, the top causality use cases, and the most exciting opportunities in the field.  

The complete show notes for this episode can be found at twimlai.com/go/616.

Today we’re joined by Dimitris Zermas, a principal scientist at agriscience company Sentera. Dimitris’ work at Sentera is focused on developing tools for precision agriculture using machine learning, including hardware like cameras and sensors, as well as ML models for analyzing the vast amount of data they acquire. We explore some specific use cases for machine learning, including plant counting, the challenges of working with classical computer vision techniques, database management, and data annotation. We also discuss their use of approaches like zero-shot learning and how they’ve taken advantage of a data-centric mindset when building a better, more cost-efficient product.

Today we’re joined by Anima Anandkumar, Bren Professor of Computing And Mathematical Sciences at Caltech and Sr Director of AI Research at NVIDIA. In our conversation, we take a broad look at the emerging field of AI for Science, focusing on both practical applications and longer-term research areas. We discuss the latest developments in the area of protein folding, and how much it has evolved since we first discussed it on the podcast in 2018, the impact of generative models and stable diffusion on the space, and the application of neural operators. We also explore the ways in which prediction models like weather models could be improved, how foundation models are helping to drive innovation, and finally, we dig into MineDojo, a new framework built on the popular Minecraft game for embodied agent research, which won a 2022 Outstanding Paper Award at NeurIPS. 

The complete show notes for this episode can be found at twimlai.com/go/614

Today we continue our AI Trends 2023 series joined by Sameer Singh, an associate professor in the department of computer science at UC Irvine and fellow at the Allen Institute for Artificial Intelligence (AI2). In our conversation with Sameer, we focus on the latest and greatest advancements and developments in the field of NLP, starting out with one that took the internet by storm just a few short weeks ago, ChatGPT. We also explore top themes like decomposed reasoning, causal modeling in NLP, and the need for “clean” data. We also discuss projects like HuggingFace’s BLOOM, the debacle that was the Galactica demo, the impending intersection of LLMs and search, use cases like Copilot, and of course, we get Sameer’s predictions for what will happen this year in the field.

The complete show notes for this episode can be found at twimlai.com/go/613.

Today we’re taking a deep dive into the latest and greatest in the world of Reinforcement Learning with our friend Sergey Levine, an associate professor, at UC Berkeley. In our conversation with Sergey, we explore some game-changing developments in the field including the release of ChatGPT and the onset of RLHF. We also explore more broadly the intersection of RL and language models, as well as advancements in offline RL and pre-training for robotics models, inverse RL, Q learning, and a host of papers along the way. Finally, you don’t want to miss Sergey’s predictions for the top developments of the year 2023! 

The complete show notes for this episode can be found at twimlai.com/go/612

Today we conclude our coverage of the 2022 NeurIPS series joined by Catherine Nakalembe, an associate research professor at the University of Maryland, and Africa Program Director under NASA Harvest. In our conversation with Catherine, we take a deep dive into her talk from the ML in the Physical Sciences workshop, Supporting Food Security in Africa using Machine Learning and Earth Observations. We discuss the broad challenges associated with food insecurity, as well as Catherine’s role and the priorities of Harvest Africa, a program focused on advancing innovative satellite-driven methods to produce automated within-season crop type and crop-specific condition products that support agricultural assessments. We explore some of the technical challenges of her work, including the limited, but growing, access to remote sensing and earth observation datasets and how the availability of that data has changed in recent years, the lack of benchmarks for the tasks she’s working on, examples of how they’ve applied techniques like multi-task learning and task-informed meta-learning, and much more. 

The complete show notes for this episode can be found at twimlai.com/go/611.

Today we conclude our AWS re:Invent 2022 series joined by Michael Kearns, a professor in the department of computer and information science at UPenn, as well as an Amazon Scholar. In our conversation, we briefly explore Michael’s broader research interests in responsible AI and ML governance and his role at Amazon. We then discuss the announcement of service cards, and their take on “model cards” at a holistic, system level as opposed to an individual model level. We walk through the information represented on the cards, as well as explore the decision-making process around specific information being omitted from the cards. We also get Michael’s take on the years-old debate of algorithmic bias vs dataset bias, what some of the current issues are around this topic, and what research he has seen (and hopes to see) addressing issues of “fairness” in large language models.  

The complete show notes for this episode can be found at twimlai.com/go/610.

Today we continue our NeurIPS 2022 series joined by Tony Jebara, VP of engineering and head of machine learning at Spotify. In our conversation with Tony, we discuss his role at Spotify and how the company’s use of machine learning has evolved over the last few years, and the business value of machine learning, specifically recommendations, hold at the company.

We dig into his talk on the intersection of reinforcement learning and lifetime value (LTV) at Spotify, which explores the application of Offline RL for user experience personalization. We discuss the various papers presented in the talk, and how they all map toward determining and increasing a user’s LTV. 

The complete show notes for this episode can be found at twimlai.com/go/609.

More than any system before it, ChatGPT has tapped into our enduring fascination with artificial intelligence, raising in a more concrete and present way important questions and fears about what AI is capable of and how it will impact us as humans. One of the concerns most frequently voiced, whether sincerely or cloaked in jest, is how ChatGPT or systems like it, will impact our livelihoods. In other words, “will ChatGPT put me out of a job???” In this episode of the podcast, I seek to answer this very question by conducting an interview in which ChatGPT is asking all the questions. (The questions are answered by a second ChatGPT, as in my own recent Interview with it, Exploring Large Laguage Models with ChatGPT.) In addition to the straight dialogue, I include my own commentary along the way and conclude with a discussion of the results of the experiment, that is, whether I think ChatGPT will be taking my job as your host anytime soon. Ultimately, though, I hope you’ll be the judge of that and share your thoughts on how ChatGPT did at my job via a comment below or on social media.

Today we continue our re:Invent 2022 series joined by Kumar Chellapilla, a general manager of ML and AI Services at AWS. We had the opportunity to speak with Kumar after announcing their recent addition of geospatial data to the SageMaker Platform. In our conversation, we explore Kumar’s role as the GM for a diverse array of SageMaker services, what has changed in the geospatial data landscape over the last 10 years, and why Amazon decided now was the right time to invest in geospatial data. We discuss the challenges of accessing and working with this data and the pain points they’re trying to solve. Finally, Kumar walks us through a few customer use cases, describes how this addition will make users more effective than they currently are, and shares his thoughts on the future of this space over the next 2-5 years, including the potential intersection of geospatial data and stable diffusion/generative models.

The complete show notes for this episode can be found at twimlai.com/go/607

Today we’re joined by Disha Singla, a senior director of machine learning engineering at Capital One. In our conversation with Disha, we explore her role as the leader of the Data Insights team at Capital One, where they’ve been tasked with creating reusable libraries, components, and workflows to make ML usable broadly across the company, as well as a platform to make it all accessible and to drive meaningful insights. We discuss the construction of her team, as well as the types of interactions and requests they receive from their customers (data scientists), productionized use cases from the platform, and their efforts to transition from batch to real-time deployment. Disha also shares her thoughts on the ROI of machine learning and getting buy-in from executives, how she sees machine learning evolving at the company over the next 10 years, and much more!

The complete show notes for this episode can be found at twimlai.com/go/606

Today we’re excited to kick off our coverage of the 2022 NeurIPS conference with Johann Brehmer, a research scientist at Qualcomm AI Research in Amsterdam. We begin our conversation discussing some of the broader problems that causality will help us solve, before turning our focus to Johann’s paper Weakly supervised causal representation learning, which seeks to prove that high-level causal representations are identifiable in weakly supervised settings. We also discuss a few other papers that the team at Qualcomm presented, including neural topological ordering for computation graphs, as well as some of the demos they showcased, which we’ll link to on the show notes page. 

The complete show notes for this episode can be found at twimlai.com/go/605.

Today we’re excited to kick off our 2022 AWS re:Invent series with a conversation with Emad Mostaque, Founder and CEO of Stability.ai. Stability.ai is a very popular name in the generative AI space at the moment, having taken the internet by storm with the release of its stable diffusion model just a few months ago. In our conversation with Emad, we discuss the story behind Stability's inception, the model's speed and scale, and the connection between stable diffusion and programming. We explore some of the spaces that Emad anticipates being disrupted by this technology, his thoughts on the open-source vs API debate, how they’re dealing with issues of user safety and artist attribution, and of course, what infrastructure they’re using to stand the model up.

The complete show notes for this episode can be found at https://twimlai.com/go/604.

Today we're joined by ChatGPT, the latest and coolest large language model developed by OpenAl. In our conversation with ChatGPT, we discuss the background and capabilities of large language models, the potential applications of these models, and some of the technical challenges and open questions in the field. We also explore the role of supervised learning in creating ChatGPT, and the use of PPO in training the model. Finally, we discuss the risks of misuse of large language models, and the best resources for learning more about these models and their applications. Join us for a fascinating conversation with ChatGPT, and learn more about the exciting world of large language models.

The complete show notes for this episode can be found at https://twimlai.com/go/603

Are AI-generating algorithms the path to artificial general intelligence(AGI)? 

Today we’re joined by Jeff Clune, an associate professor of computer science at the University of British Columbia, and faculty member at the Vector Institute. In our conversation with Jeff, we discuss the broad ambitious goal of the AI field, artificial general intelligence, where we are on the path to achieving it, and his opinion on what we should be doing to get there, specifically, focusing on AI generating algorithms. With the goal of creating open-ended algorithms that can learn forever, Jeff shares his three pillars to an AI-GA, meta-learning architectures, meta-learning algorithms, and auto-generating learning environments. Finally, we discuss the inherent safety issues with these learning algorithms and Jeff’s thoughts on how to combat them, and what the not-so-distant future holds for this area of research. 

The complete show notes for this episode can be found at twimlai.com/go/602.

Today we’re joined by Cedric Cocaud, the chief engineer of the Wayfinder Group at Acubed, the innovation center for aircraft manufacturer Airbus. In our conversation with Cedric, we explore some of the technical challenges of innovation in the aircraft space, including autonomy. Cedric’s work on Project Vahana, Acubed’s foray into air taxis, attempted to leverage work in the self-driving car industry to develop fully autonomous planes. We discuss some of the algorithms being developed for this work, the data collection process, and Cedric’s thoughts on using synthetic data for these tasks. We also discuss the challenges of labeling the data, including programmatic and automated labeling, and much more.

Today we’re joined by Heather Nolis, a principal machine learning engineer at T-Mobile. In our conversation with Heather, we explored her machine learning journey at T-Mobile, including their initial proof of concept project, which held the goal of putting their first real-time deep learning model into production. We discuss the use case, which aimed to build a model customer intent model that would pull relevant information about a customer during conversations with customer support. This process has now become widely known as blank assist. We also discuss the decision to use supervised learning to solve this problem and the challenges they faced when developing a taxonomy. Finally, we explore the idea of using small models vs uber-large models, the hardware being used to stand up their infrastructure, and how Heather thinks about the age-old question of build vs buy. 

Today we’re joined by return guest Ken Goldberg, a professor at UC Berkeley and the chief scientist at Ambi Robotics. It’s been a few years since our initial conversation with Ken, so we spent a bit of time talking through the progress that has been made in robotics in the time that has passed. We discuss Ken’s recent work, including the paper Autonomously Untangling Long Cables, which won Best Systems Paper at the RSS conference earlier this year, including the complexity of the problem and why it is classified as a systems challenge, as well as the advancements in hardware that made solving this problem possible. We also explore Ken’s thoughts on the push towards simulation by research entities and large tech companies, and the potential for causal modeling to find its way into robotics. Finally, we discuss the recent showcase of Optimus, Tesla, and Elon Musk’s “humanoid” robot and how far we are from it being a viable piece of technology.

The complete show notes for this episode can be found at twimlai.com/go/599.

Today friend of the show and esteemed guest host John Bohannon is back with another great interview, this time around joined by Oren Etzioni, former CEO of the Allen Institute for AI, where he is currently an advisor. In our conversation with Oren, we discuss his philosophy as a researcher and how that has manifested in his pivot to institution builder. We also explore his thoughts on the current landscape of NLP, including the emergence of LLMs and the hype being built up around AI systems from folks like Elon Musk. Finally, we explore some of the research coming out of AI2, including Semantic Scholar, an AI-powered research tool analogous to arxiv, and the somewhat controversial Delphi project, a research prototype designed to model people’s moral judgments on a variety of everyday situations.

Over the last few years, it’s been established that your ML team needs at least some basic tooling in order to be effective, providing support for various aspects of the machine learning workflow, from data acquisition and management, to model development and optimization, to model deployment and monitoring.

But how do you get there? Many tools available off the shelf, both commercial and open source, can help.

At the extremes, these tools can fall into one of a couple of buckets. End-to-end platforms that try to provide support for many aspects of the ML lifecycle, and specialized tools that offer deep functionality in a particular domain or area.

At TWIMLcon: AI Platforms 2022, our panelists debated the merits of these approaches in The Great MLOps Debate: End-to-End ML Platforms vs Specialized Tools.

Much of the way we talk and think about MLOps comes from the perspective of large consumer internet companies like Facebook or Google. If you work at a FAANG company, these approaches might work well for you. But what about if you work at one of the many small, B2B companies that stand to benefit through the use of machine learning? How should you be thinking about MLOps and the ML lifecycle in that case? In this live podcast interview from TWIMLcon: AI Platforms 2022, Sam Charrington explores these questions with Jacopo Tagliabue, whose perspectives and contributions on scaling down MLOps have served to make the field more accessible and relevant to a wider array of practitioners.

Today we’re joined by Ali Rodell, a senior director of machine learning engineering at Capital One. In our conversation with Ali, we explore his role as the head of model development platforms at Capital One, including how his 25+ years in software development have shaped his view on building platforms and the evolution of the platforms space over the last 10 years. We discuss the importance of a healthy open source tooling ecosystem, Capital One’s use of various open source capabilites like kubeflow and kubernetes to build out platforms, and some of the challenges that come along with modifying/customizing these tools to work for him and his teams. Finally, we explore the range of user personas that need to be accounted for when making decisions about tooling, supporting things like Jupyter notebooks and other low level tools, and how that can be potentially challenging in a highly regulated environment like the financial industry.

The complete show notes for this episode can be found at twimlai.com/go/595

Today we’re joined by Vasi Philomin, vice president of AI services at AWS, joins us for our first in-person interview since 2019! In our conversation with Vasi, we discussed the recently released Amazon Code Whisperer, a developer-focused coding companion. We begin by exploring Vasi’s role and the various products under the banner of cognitive and non-cognitive services, and how those came together where Code Whisperer fits into the equation and some of the differences between Code Whisperer and some of the other recently released coding companions like GitHub Copilot. We also discuss the training corpus for the model, and how they’ve dealt with the potential issues of bias that arise when training LLMs with crawled web data, and Vasi’s thoughts on what the path of innovation looks like for Code Whisperer. 

At the end of our conversation, Vasi was gracious enough to share a quick live demo of Code Whisperer, so you can catch that here.

TWIMLcon: AI Platforms 2022 is just a day away! If you're interested in all things MLOps and Platforms/Infrastructure technology, this is the event for you! Register now at https://twimlcon.com/attend for FREE!

Today we’re joined by Vidyut Naware, the director of machine learning and artificial intelligence at Paypal. As the leader of the ML/AI organization at Paypal, Vidyut is responsible for all things applied, from R&D to MLOps infrastructure. In our conversation, we explore the work being done in four major categories, hardware/compute, data, applied responsible AI, and tools, frameworks, and platforms. We also discuss their use of federated learning and delayed supervision models for use cases like anomaly detection and fraud prevention, research into quantum computing and causal inference, as well as applied use cases like graph machine learning and collusion detection. 

The complete show notes for this episode can be found at twimlai.com/go/593

Today we’re back with another installment of our Data-Centric AI series, joined by Wendy Foster, a director of engineering & data science at Shopify. In our conversation with Wendy, we explore the differences between data-centric and model-centric approaches and how they manifest at Shopify, including on her team, which is responsible for utilizing merchant and product data to assist individual vendors on the platform. We discuss how they address, maintain, and improve data quality, emphasizing the importance of coverage and “freshness” data when solving constantly evolving use cases. Finally, we discuss how data is taxonomized at the company and the challenges that present themselves when producing large-scale ML models, future use cases that Wendy expects her team to tackle, and we briefly explore Merlin, Shopify’s new ML platform (that you can hear more about at TWIMLcon!), and how it fits into the broader scope of ML at the company.

The complete show notes for this episode can be found at twimlai.com/go/592

Today we’re joined by Bayan Bruss, a Sr. director of applied ML research at Capital One. In our conversation with Bayan, we dig into his work in applying various deep learning techniques to tabular data, including taking advancements made in other areas like graph CNNs and other traditional graph mining algorithms and applying them to financial services applications. We discuss why despite a “flood” of innovation in the field, work on tabular data doesn’t elicit as much fanfare despite its broad use across businesses, Bayan’s experience with the difficulty of making deep learning work on tabular data, and what opportunities have been presented for the field with the emergence of multi-modality and transformer models. We also explore a pair of papers from Bayan’s team, focused on both transformers and transfer learning for tabular data. 

The complete show notes for this episode can be found at twimlai.com/go/591

Today we’re joined by Orit Peleg, an assistant professor at the University of Colorado, Boulder. Orit’s work focuses on understanding the behavior of disordered living systems, by merging tools from physics, biology, engineering, and computer science. In our conversation, we discuss how Orit found herself exploring problems of swarming behaviors and their relationship to distributed computing system architecture and spiking neurons. We look at two specific areas of research, the first focused on the patterns observed in firefly species, how the data is collected, and the types of algorithms used for optimization. Finally, we look at how Orit’s research with fireflies translates to a completely different insect, the honeybee, and what the next steps are for investigating these and other insect families.

The complete show notes for this episode can be found at twimlai.com/go/590

In this extra special episode of the TWIML AI Podcast, a friend of the show John Bohannon leads a jam-packed conversation with Hugging Face’s recently appointed head of research Douwe Kiela. In our conversation with Douwe, we explore his role at the company, how his perception of Hugging Face has changed since joining, and what research entails at the company. We discuss the emergence of the transformer model and the emergence of BERT-ology, the recent shift to solving more multimodal problems, the importance of this subfield as one of the “Grand Directions'' of Hugging Face’s research agenda, and the importance of BLOOM, the open-access Multilingual Language Model that was the output of the BigScience project. Finally, we get into how Douwe’s background in philosophy shapes his view of current projects, as well as his projections for the future of NLP and multimodal ML.

The complete show notes for this episode can be found at twimlai.com/go/589

Today we’re joined by Bill Vass, a VP of engineering at Amazon Web Services. Bill spoke at the most recent AWS re:MARS conference, where he delivered an engineering Keynote focused on some recent updates to Amazon sagemaker, including its support for synthetic data generation. In our conversation, we discussed all things synthetic data, including the importance of data quality when creating synthetic data, and some of the use cases that this data is being created for, including warehouses and in the case of one of their more recent acquisitions, iRobot, synthetic house generation. We also explore Astro, the household robot for home monitoring, including the types of models running it, is running, what type of on-device sensor suite it has, the relationship between the robot and the cloud, and the role of simulation. 

The complete show notes for this episode can be found at twimlai.com/go/588

Today we’re joined by Jeff Gehlhaar, vice president of technology at Qualcomm Technologies. In our annual conversation with Jeff, we dig into the relationship between Jeff’s team on the product side and the research team, many of whom we’ve had on the podcast over the last few years. We discuss the challenges of real-world neural network deployment and doing quantization on-device, as well as a look at the tools that power their AI Stack. We also explore a few interesting automotive use cases, including automated driver assistance, and what advancements Jeff is looking forward to seeing in the next year.

The complete show notes for this episode can be found at twimlai.com/go/587

Today we close out our ICML 2022 coverage joined by Sharad Goel, a professor of public policy at Harvard University. In our conversation with Sharad, we discuss his Outstanding Paper award winner Causal Conceptions of Fairness and their Consequences, which seeks to understand what it means to apply causality to the idea of fairness in ML. We explore the two broad classes of intent that have been conceptualized under the subfield of causal fairness and how they differ, the distinct ways causality is treated in economic and statistical contexts vs a computer science and algorithmic context, and why policies are created in the context of causal definitions are suboptimal broadly.

The complete show notes for this episode can be found at twimlai.com/go/586

Today we continue our ICML coverage joined by Melika Payvand, a research scientist at the Institute of Neuroinformatics at the University of Zurich and ETH Zurich. Melika spoke at the Hardware Aware Efficient Training (HAET) Workshop, delivering a keynote on Brain-inspired hardware and algorithm co-design for low power online training on the edge. In our conversation with Melika, we explore her work at the intersection of ML and neuroinformatics, what makes the proposed architecture “brain-inspired”, and how techniques like online learning fit into the picture. We also discuss the characteristics of the devices that are running the algorithms she’s creating, and the challenges of adapting online learning-style algorithms to this hardware.

The complete show notes for this episode can be found at twimlai.com/go/585

Today we’re joined by Arash Behboodi, a machine learning researcher at Qualcomm Technologies. In our conversation with Arash, we explore his paper Equivariant Priors for Compressed Sensing with Unknown Orientation, which proposes using equivariant generative models as a prior means to show that signals with unknown orientations can be recovered with iterative gradient descent on the latent space of these models and provide additional theoretical recovery guarantees. We discuss the differences between compression and compressed sensing, how he was able to evolve a traditional VAE architecture to understand equivalence, and some of the research areas he’s applying this work, including cryo-electron microscopy. We also discuss a few of the other papers that his colleagues have submitted to the conference, including Overcoming Oscillations in Quantization-Aware Training, Variational On-the-Fly Personalization, and CITRIS: Causal Identifiability from Temporal Intervened Sequences.

The complete show notes for this episode can be found at twimlai.com/go/584

Today we continue our Data-Centric AI Series joined by Audrey Smith, the COO at MLtwist, and a recent participant in our panel on DCAI. In our conversation, we do a deep dive into data labeling for ML, exploring the typical journey for an organization to get started with labeling, her experience when making decisions around in-house vs outsourced labeling, and what commitments need to be made to achieve high-quality labels. We discuss how organizations that have made significant investments in labelops typically function, how someone working on an in-house labeling team approaches new projects, the ethical considerations that need to be taken for remote labeling workforces, and much more!

The complete show notes for this episode can be found at twimlai.com/go/583

Today we’re joined by Richard Socher, the CEO of You.com. In our conversation with Richard, we explore the inspiration and motivation behind the You.com search engine, and how it differs from the traditional google search engine experience. We discuss some of the various ways that machine learning is used across the platform including how they surface relevant search results and some of the recent additions like code completion and a text generator that can write complete essays and blog posts. Finally, we talk through some of the projects we covered in our last conversation with Richard, namely his work on Salesforce’s AI Economist project. 

The complete show notes for this episode can be found at twimlai.com/go/582

Today we wrap up our coverage of the 2022 CVPR conference joined by Aljosa Osep, a postdoc at the Technical University of Munich & Carnegie Mellon University. In our conversation with Aljosa, we explore his broader research interests in achieving robot vision, and his vision for what it will look like when that goal is achieved. The first paper we dig into is Text2Pos: Text-to-Point-Cloud Cross-Modal Localization, which proposes a cross-modal localization module that learns to align textual descriptions with localization cues in a coarse-to-fine manner. Next up, we explore the paper Forecasting from LiDAR via Future Object Detection, which proposes an end-to-end approach for detection and motion forecasting based on raw sensor measurement as opposed to ground truth tracks. Finally, we discuss Aljosa’s third and final paper Opening up Open-World Tracking, which proposes a new benchmark to analyze existing efforts in multi-object tracking and constructs a baseline for these tasks.

The complete show notes for this episode can be found at twimlai.com/go/581

Today we continue our CVPR series joined by Kate Saenko, an associate professor at Boston University and a consulting professor for the MIT-IBM Watson AI Lab. In our conversation with Kate, we explore her research in multimodal learning, which she spoke about at the Multimodal Learning and Applications Workshop, one of a whopping 6 workshops she spoke at. We discuss the emergence of multimodal learning, the current research frontier, and Kate’s thoughts on the inherent bias in LLMs and how to deal with it. We also talk through some of the challenges that come up when building out applications, including the cost of labeling, and some of the methods she’s had success with. Finally, we discuss Kate’s perspective on the monopolizing of computing resources for “foundational” models, and her paper Unsupervised Domain Generalization by learning a Bridge Across Domains.

The complete show notes for this episode can be found at twimlai.com/go/580

Today we kick off our annual coverage of the CVPR conference joined by Fatih Porikli, Senior Director of Engineering at Qualcomm AI Research. In our conversation with Fatih, we explore a trio of CVPR-accepted papers, as well as a pair of upcoming workshops at the event. The first paper, Panoptic, Instance and Semantic Relations: A Relational Context Encoder to Enhance Panoptic Segmentation, presents a novel framework to integrate semantic and instance contexts for panoptic segmentation. Next up, we discuss Imposing Consistency for Optical Flow Estimation, a paper that introduces novel and effective consistency strategies for optical flow estimation. The final paper we discuss is IRISformer: Dense Vision Transformers for Single-Image Inverse Rendering in Indoor Scenes, which proposes a transformer architecture to simultaneously estimate depths, normals, spatially-varying albedo, roughness, and lighting from a single image of an indoor scene. For each paper, we explore the motivations and challenges and get concrete examples to demonstrate each problem and solution presented.

The complete show notes for this episode can be found at twimlai.com/go/579

Today we’re joined by Adam Wood, Director of Data Governance and Data Quality at Mastercard. In our conversation with Adam, we explore the challenges that come along with data governance at a global scale, including dealing with regional regulations like GDPR and federating records at scale. We discuss the role of feature stores in keeping track of data lineage and how Adam and his team have dealt with the challenges of metadata management, how large organizations like Mastercard are dealing with enabling feature reuse, and the steps they take to alleviate bias, especially in scenarios like acquisitions. Finally, we explore data quality for data science and why Adam sees it as an encouraging area of growth within the company, as well as the investments they’ve made in tooling around data management, catalog, feature management, and more.

The complete show notes for this episode can be found at twimlai.com/go/578

In the latest installment of our Data-Centric AI series, we’re joined by a friend of the show Mike Del Balso, Co-founder and CEO of Tecton. If you’ve heard any of our other conversations with Mike, you know we spend a lot of time discussing feature stores, or as he now refers to them, feature platforms. We explore the current complexity of data infrastructure broadly and how that has changed over the last five years, as well as the maturation of streaming data platforms. We discuss the wide vs deep paradox that exists around ML tooling, and the idea around the “ML Flywheel”, a strategy that leverages data to accelerate machine learning. Finally, we spend time discussing internal ML team construction, some of the challenges that organizations face when building their ML platforms teams, and how they can avoid the pitfalls as they arise.

The complete show notes for this episode can be found at twimlai.com/go/577

Today we continue our Data-centric AI series joined by Shayan Mohanty, CEO at Watchful. In our conversation with Shayan, we focus on the data labeling aspect of the machine learning process, and ways that a data-centric approach could add value and reduce cost by multiple orders of magnitude. Shayan helps us define “data-centric”, while discussing the main challenges that organizations face when dealing with labeling, how these problems are currently being solved, and how techniques like active learning and weak supervision could be used to more effectively label. We also explore the idea of machine teaching, which focuses on using techniques that make the model training process more efficient, and what organizations need to be successful when trying to make the aforementioned mindset shift to DCAI. 

The complete show notes for this episode can be found at twimlai.com/go/576

This week, we continue our conversations around the topic of Data-Centric AI joined by a friend of the show Adrien Gaidon, the head of ML research at the Toyota Research Institute (TRI). In our chat, Adrien expresses a fourth, somewhat contrarian, viewpoint to the three prominent schools of thought that organizations tend to fall into, as well as a great story about how the breakthrough came via an unlikely source. We explore his principle-centric approach to machine learning as well as the role of self-supervised machine learning and synthetic data in this and other research threads. Make sure you’re following along with the entire DCAI series at twimlai.com/go/dcai.

The complete show notes for this episode can be found at twimlai.com/go/575

Today we kick things off with a conversation with D. Sculley, a director on the Google Brain team. Many listeners of today’s show will know D. from his work on the paper, The Hidden Technical Debt in Machine Learning Systems, and of course, the infamous diagram. D. has recently translated the idea of technical debt into data debt, something we spend a bit of time on in the interview.

We discuss his view of the concept of DCAI, where debt fits into the conversation of data quality, and what a shift towards data-centrism looks like in a world of increasingly larger models i.e. GPT-3 and the recent PALM models. We also explore common sources of data debt, what are things that the community can and have done to mitigate these issues, the usefulness of causal inference graphs in this work, and much more! If you enjoyed this interview or want to hear more on this topic, check back on the DCAI series page weekly at https://twimlai.com/podcast/twimlai/series/data-centric-ai.

The complete show notes for this episode can be found at twimlai.com/go/574

Today we’re joined by Rob Walker, VP of decisioning & analytics and gm of one-to-one customer engagement at Pegasystems. Rob, who you might know from his previous appearances on the podcast, joins us to discuss his work on AI and ML in the context of customer engagement and decisioning, the various problems that need to be solved, including solving the “next best” problem. We explore the distinction between the idea of the next best action and determining it from a recommender system, how the combination of machine learning and heuristics are currently co-existing in engagements, scaling model evaluation, and some of the challenges they’re facing when dealing with problems of responsible AI and how they’re managed. Finally, we spend a few minutes digging into the upcoming PegaWorld conference, and what attendees should anticipate at the event.

The complete show notes for this episode can be found at twimlai.com/go/573

Today we close out our coverage of the ICLR series joined by Meg Mitchell, chief ethics scientist and researcher at Hugging Face. In our conversation with Meg, we discuss her participation in the WikiM3L Workshop, as well as her transition into her new role at Hugging Face, which has afforded her the ability to prioritize coding in her work around AI ethics. We explore her thoughts on the work happening in the fields of data curation and data governance, her interest in the inclusive sharing of datasets and creation of models that don't disproportionately underperform or exploit subpopulations, and how data collection practices have changed over the years. 

We also touch on changes to data protection laws happening in some pretty uncertain places, the evolution of her work on Model Cards, and how she’s using this and recent Data Cards work to lower the barrier to entry to responsibly informed development of data and sharing of data.

The complete show notes for this episode can be found at twimlai.com/go/572

Today we continue our ICLR coverage joined by Been Kim, a staff research scientist at Google Brain, and an ICLR 2022 Invited Speaker. Been, whose research has historically been focused on interpretability in machine learning, delivered the keynote Beyond interpretability: developing a language to shape our relationships with AI, which explores the need to study AI machines as scientific objects, in isolation and with humans, which will provide principles for tools, but also is necessary to take our working relationship with AI to the next level. 

Before we dig into Been’s talk, she characterizes where we are as an industry and community with interpretability, and what the current state of the art is for interpretability techniques. We explore how the Gestalt principles appear in neural networks, Been’s choice to characterize communication with machines as a language as opposed to a set of principles or foundational understanding, and much much more.

The complete show notes for this episode can be found at twimlai.com/go/571

Today we’re joined by Auke Wiggers, an AI research scientist at Qualcomm. In our conversation with Auke, we discuss his team’s recent research on data compression using generative models. We discuss the relationship between historical compression research and the current trend of neural compression, and the benefit of neural codecs, which learn to compress data from examples. We also explore the performance evaluation process and the recent developments that show that these models can operate in real-time on a mobile device. Finally, we discuss another ICLR paper, “Transformer-based transform coding”, that proposes a vision transformer-based architecture for image and video coding, and some of his team’s other accepted works at the conference. 

The complete show notes for this episode can be found at twimlai.com/go/570

Today we’re joined by Irwan Bello, formerly a research scientist at Google Brain, and now on the founding team at a stealth AI startup. We begin our conversation with an exploration of Irwan’s recent paper, Designing Effective Sparse Expert Models, which acts as a design guide for building sparse large language model architectures. We discuss mixture of experts as a technique, the scalability of this method, and it's applicability beyond NLP tasks the data sets this experiment was benchmarked against. We also explore Irwan’s interest in the research areas of alignment and retrieval, talking through interesting lines of work for each area including instruction tuning and direct alignment.

The complete show notes for this episode can be found at twimlai.com/go/569

Today we’re joined by friend of the show Timnit Gebru, the founder and executive director of DAIR, the Distributed Artificial Intelligence Research Institute. In our conversation with Timnit, we discuss her journey to create DAIR, their goals and some of the challenges shes faced along the way. We start is the obvious place, Timnit being “resignated” from Google after writing and publishing a paper detailing the dangers of large language models, the fallout from that paper and her firing, and the eventual founding of DAIR. We discuss the importance of the “distributed” nature of the institute, how they’re going about figuring out what is in scope and out of scope for the institute’s research charter, and what building an institution means to her. We also explore the importance of independent alternatives to traditional research structures, if we should be pessimistic about the impact of internal ethics and responsible AI teams in industry due to the overwhelming power they wield, examples she looks to of what not to do when building out the institute, and much much more!

The complete show notes for this episode can be found at twimlai.com/go/568

Today we’re joined by Doina Precup, a research team lead at DeepMind Montreal, and a professor at McGill University. In our conversation with Doina, we discuss her recent research interests, including her work in hierarchical reinforcement learning, with the goal being agents learning abstract representations, especially over time. We also explore her work on reward specification for RL agents, where she hypothesizes that a reward signal in a complex environment could lead an agent to develop attributes of intuitive intelligence. We also dig into quite a few of her papers, including On the Expressivity of Markov Reward, which won a NeruIPS 2021 outstanding paper award. Finally, we discuss the analogy between hierarchical RL and CNNs, her work in continual RL, and her thoughts on the evolution of RL in the recent past and present, and the biggest challenges facing the field going forward.

The complete show notes for this episode can be found at twimlai.com/go/567

Today we’re joined by Bharath Ramsundar, founder and CEO of Deep Forest Sciences. In our conversation with Bharath, we explore his work on the DeepChem, an open-source library for drug discovery, materials science, quantum chemistry, and biology tools. We discuss the challenges that biotech and pharmaceutical companies are facing as they attempt to incorporate AI into the drug discovery process, where the innovation frontier is, and what the promise is for AI in this field in the near term. We also dig into the origins of DeepChem and the problems it's solving for practitioners, the capabilities that are enabled when using this library as opposed to others, and MoleculeNET, a dataset and benchmark focused on molecular design that lives within the DeepChem suite.

The complete show notes for this episode can be found at twimlai.com/go/566

Today we’re joined by Sebastian Raschka, an assistant professor at the University of Wisconsin-Madison and lead AI educator at Grid.ai. In our conversation with Sebastian, we explore his work around AI education, including the “hands-on” philosophy that he takes when building these courses, his recent book Machine Learning with PyTorch and Scikit-Learn, his advise to beginners in the field when they’re trying to choose tools and frameworks, and more. 

We also discuss his work on Pytorch Lightning, a platform that allows users to organize their code and integrate it into other technologies, before switching gears and discuss his recent research efforts around ordinal regression, including a ton of great references that we’ll link on the show notes page below! 

The complete show notes for this episode can be found at twimlai.com/go/565

Today we’re joined by Thomas Wolf, co-founder and chief science officer at Hugging Face 🤗. We cover a ton of ground In our conversation, starting with Thomas’ interesting backstory as a quantum physicist and patent lawyer, and how that lead him to a career in machine learning. We explore how Hugging Face began, what the current direction is for the company, and how much of their focus is NLP and language models versus other disciplines. We also discuss the BigScience project, a year-long research workshop where 1000+ researchers of all backgrounds and disciplines have come together to create an 800GB multilingual dataset and model. We talk through their approach to curating the dataset, model evaluation at this scale, and how they differentiate their work from projects like Eluther AI. Finally, we dig into Thomas’ work on multimodality, his thoughts on the metaverse, his new book NLP with Transformers, and much more!

The complete show notes for this episode can be found at twimlai.com/go/564

Today we’re joined by Murali Akula, a Sr. director of Software Engineering at Qualcomm. In our conversation with Murali, we explore his role at Qualcomm, where he leads the corporate research team focused on the development and deployment of AI onto Snapdragon chips, their unique definition of “full stack”, and how that philosophy permeates into every step of the software development process. We explore the complexities that are unique to doing machine learning on resource constrained devices, some of the techniques that are being applied to get complex models working on mobile devices, and the process for taking these models from research into real-world applications. We also discuss a few more tools and recent developments, including DONNA for neural architecture search, X-Distill, a method of improving the self-supervised training of monocular depth, and the AI Model Effeciency Toolkit, a library that provides advanced quantization and compression techniques for trained neural network models.

The complete show notes for this episode can be found at twimlai.com/go/563

Today we’re joined by Subutai Ahmad, VP of research at Numenta. While we’ve had numerous conversations about the biological inspirations of deep learning models with folks working at the intersection of deep learning and neuroscience, we dig into uncharted territory with Subutai. We set the stage by digging into some of fundamental ideas behind Numenta’s research and the present landscape of neuroscience, before exploring our first big topic of the podcast: the cortical column. Cortical columns are a group of neurons in the cortex of the brain which have nearly identical receptive fields; we discuss the behavior of these columns, why they’re a structure worth mimicing computationally, how far along we are in understanding the cortical column, and how these columns relate to neurons.

We also discuss what it means for a model to have inherent 3d understanding and for computational models to be inherently sensory motor, and where we are with these lines of research. Finally, we dig into our other big idea, sparsity. We explore the fundamental ideals of sparsity and the differences between sparse and dense networks, and applying sparsity and optimization to drive greater efficiency in current deep learning networks, including transformers and other large language models. 

The complete show notes for this episode can be found at twimlai.com/go/562

Today we’re joined by Jennifer Glore, VP of customer engineering at SambaNova Systems. In our conversation with Jennifer, we discuss how, and why, Sambanova, who is primarily focused on building hardware to support machine learning applications, has built a GPT language model for the financial services industry. Jennifer shares her thoughts on the progress of industries like banking and finance, as well as other traditional organizations, in their attempts at using transformers and other models, and where they’ve begun to see success, as well as some of the hidden challenges that orgs run into that impede their progress. Finally, we explore their experience replicating the GPT-3 paper from a R&D perspective, how they’re addressing issues of predictability, controllability, governance, etc, and much more.

The complete show notes for this episode can be found at twimlai.com/go/561

Today we’re joined by Kamyar Azizzadenesheli, an assistant professor at Purdue University, to close out our AI Rewind 2021 series! In this conversation, we focused on all things deep reinforcement learning, starting with a general overview of the direction of the field, and though it might seem to be slowing, thats just a product of the light being shined constantly on the CV and NLP spaces. We dig into themes like the convergence of RL methodology with both robotics and control theory, as well as a few trends that Kamyar sees over the horizon, such as self-supervised learning approaches in RL. We also talk through Kamyar’s predictions for RL in 2022 and beyond. This was a fun conversation, and I encourage you to look through all the great resources that Kamyar shared on the show notes page at twimlai.com/go/560!

Today we’re joined by Rishabh Agarwal, a research scientist at Google Brain in Montreal. In our conversation with Rishabh, we discuss his recent paper Deep Reinforcement Learning at the Edge of the Statistical Precipice, which won an outstanding paper award at the most recent NeurIPS conference. In this paper, Rishabh and his coauthors call for a change in how deep RL performance is reported on benchmarks when using only a few runs, acknowledging that typically, DeepRL algorithms are evaluated by the performance on a large suite of tasks. Using the Atari 100k benchmark, they found substantial disparities in the conclusions from point estimates alone versus statistical analysis. We explore the reception of this paper from the research community, some of the more surprising results, what incentives researchers have to implement these types of changes in self-reporting when publishing, and much more.

The complete show notes for this episode can be found at twimlai.com/go/559

Today we’re joined by Rafael Gomez-Bombarelli, an assistant professor in the department of material science and engineering at MIT. In our conversation with Rafa, we explore his goal of ​​fusing machine learning and atomistic simulations for designing materials, a topic he spoke about at the recent SigOpt AI & HPC Summit. We discuss the two ways in which he thinks of material design, virtual screening and inverse design, as well as the unique challenges each technique presents. We also talk through the use of generative models for simulation, the type of training data necessary for these tasks, and if he’s building hand-coded simulations vs existing packages or tools. Finally, we explore the dynamic relationship between simulation and modeling and how the results of one drive the others efforts, and how hyperparameter optimization gets incorporated into the various projects.

The complete show notes for this episode can be found at twimlai.com/go/558

Today we’re joined by Patrick Heimbach, a professor at the University of Texas working at the intersection of ML and oceanography. In our conversation with Patrick, we explore some of the challenges of computational oceanography, the potential use cases for machine learning in this field, as well as how it can be used to support scientists in solving simulation problems, and the role of differential programming and how it is expressed in his work. 

The complete show notes for this episode can be found at twimlai.com/go/557

Today we continue our AI Rewind 2021 series joined by a friend of the show, assistant professor at Carnegie Mellon University, and AI Rewind veteran, Zack Lipton! In our conversation with Zack, we touch on recurring themes like “NLP Eating AI” and the recent slowdown in innovation in the field, the redistribution of resources across research problems, and where the opportunities for real breakthroughs lie. We also discuss problems facing the current peer-review system, notable research from last year like the introduction of the WILDS library, and the evolution of problems (and potential solutions) in fairness, bias, and equity. Of course, we explore some of the use cases and application areas that made notable progress in 2021, what Zack is looking forward to in 2022 and beyond, and much more!

The complete show notes for this episode can be found at twimlai.com/go/556

Today we’re joined by Jonathan Le Roux, a senior principal research scientist at Mitsubishi Electric Research Laboratories (MERL). At MERL, Jonathan and his team are focused on using machine learning to solve the “cocktail party problem”, focusing on not only the separation of speech from noise, but also the separation of speech from speech. In our conversation with Jonathan, we focus on his paper The Cocktail Fork Problem: Three-Stem Audio Separation For Real-World Soundtracks, which looks to separate and enhance a complex acoustic scene into three distinct categories, speech, music, and sound effects. We explore the challenges of working with such noisy data, the model architecture used to solve this problem, how ML/DL fits into solving the larger cocktail party problem, future directions for this line of research, and much more!

The complete show notes for this episode can be found at twimlai.com/go/555

Today we’re joined by Karianne Bergen, an assistant professor at Brown University. In our conversation with Karianne, we explore her work at the intersection of earthquake seismology and machine learning, where she’s working on interpretable data classification for seismology. We discuss some of the challenges that present themselves when trying to solve this problem, and the state of applying machine learning to seismological events and earth sciences. Karianne also shares her thoughts on the different relationships that computer scientists and natural scientists have with machine learning, and how to bridge that gap to create tools that work broadly for all scientists.

The complete show notes for this episode can be found at twimlai.com/go/554

Today we’re joined by Arun Kumarm, an associate professor at UC San Diego. We had the pleasure of catching up with Arun prior to the Workshop on Databases and AI at NeurIPS 2021, where he delivered the talk “The New DBfication of ML/AI.” In our conversation, we explore this “database-ification” of machine learning, a concept analogous to the transformation of relational SQL computation. We discuss the relationship between the ML and database fields and how the merging of the two could have positive outcomes for the end-to-end ML workflow, and a few tools that his team has developed, Cerebro, a tool for reproducible model selection, and SortingHat, a tool for automating data prep, and how tools like these and others affect Arun’s outlook on the future of machine learning platforms and MLOps.

The complete show notes for this episode can be found at twimlai.com/go/553

Today we’re joined by Meredith Broussard, an associate professor at NYU & research director at the NYU Alliance for Public Interest Technology. Meredith was a keynote speaker at the recent NeurIPS conference, and we had the pleasure of speaking with her to discuss her talk from the event, and her upcoming book, tentatively titled More Than A Glitch: What Everyone Needs To Know About Making Technology Anti-Racist, Accessible, And Otherwise Useful To All.

In our conversation, we explore Meredith’s work in the field of public interest technology, and her view of the relationship between technology and artificial intelligence. Meredith and Sam talk through real-world scenarios where an emphasis on monitoring bias and responsibility would positively impact outcomes, and how this type of monitoring parallels the infrastructure that many organizations are already building out. Finally, we talk through the main takeaways from Meredith’s NeurIPS talk, and how practitioners can get involved in the work of building and deploying public interest technology.

The complete show notes for this episode can be found at twimlai.com/go/552

Today we’re joined by Sebastian Bubeck a sr principal research manager at Microsoft, and author of the paper A Universal Law of Robustness via Isoperimetry, a NeurIPS 2021 Outstanding Paper Award recipient. We begin our conversation with Sebastian with a bit of a primer on convex optimization, a topic that hasn’t come up much in previous interviews. We explore the problem that convex optimization is trying to solve, the application of convex optimization to multi-armed bandit problems, metrical task systems and solving the K-server problem. We then dig into Sebastian’s paper, which looks to prove that for a broad class of data distributions and model classes, overparameterization is necessary if one wants to interpolate the data. Finally, we discussed the relationship between the paper and the work being done in the adversarial robustness community.

The complete show notes for this episode can be found at twimlai.com/go/551

Today we’re joined by friend of the show John Bohannon, the director of science at Primer AI, to help us showcase all of the great achievements and accomplishments in NLP in 2021! In our conversation, John shares his two major takeaways from last year, 1) NLP as we know it has changed, and we’re back into the incremental phase of the science, and 2) NLP is “eating” the rest of machine learning. We explore the implications of these two major themes across the discipline, as well as best papers, up and coming startups, great things that did happen, and even a few bad things that didn’t. Finally, we explore what 2022 and beyond will look like for NLP, from multilingual NLP to use cases for the influx of large auto-regressive language models like GPT-3 and others, as well as ethical implications that are reverberating across domains and the changes that have been ushered in in that vein.

The complete show notes for this episode can be found at twimlai.com/go/550

Happy New Year! We’re excited to kick off 2022 joined by Georgia Gkioxari, a research scientist at Meta AI, to showcase the best advances in the field of computer vision over the past 12 months, and what the future holds for this domain. 

Welcome back to AI Rewind!

In our conversation Georgia highlights the emergence of the transformer model in CV research, what kind of performance results we’re seeing vs CNNs, and the immediate impact of NeRF, amongst a host of other great research. We also explore what is ImageNet’s place in the current landscape, and if it's time to make big changes to push the boundaries of what is possible with image, video and even 3D data, with challenges like the Metaverse, amongst others, on the horizon. Finally, we touch on the startups to keep an eye on, the collaborative efforts of software and hardware researchers, and the vibe of the “ImageNet moment” being upon us once again.

The complete show notes for this episode can be found at twimlai.com/go/549

Today we close out the 2021 NeurIPS series joined by Alison Gopnik, a professor at UC Berkeley and an invited speaker at the Causal Inference & Machine Learning: Why now? Workshop. In our conversation with Alison, we explore the question, “how is it that we can know so much about the world around us from so little information?,” and how her background in psychology, philosophy, and epistemology has guided her along the path to finding this answer through the actions of children. We discuss the role of causality as a means to extract representations of the world and how the “theory theory” came about, and how it was demonstrated to have merit. We also explore the complexity of causal relationships that children are able to deal with and what that can tell us about our current ML models, how the training and inference stages of the ML lifecycle are akin to childhood and adulthood, and much more!

The complete show notes for this episode can be found at twimlai.com/go/548

Today we continue our NeurIPS coverage joined by Tina Eliassi-Rad, a professor at Northeastern University, and an invited speaker at the I Still Can't Believe It's Not Better! Workshop. In our conversation with Tina, we explore her research at the intersection of network science, complex networks, and machine learning, how graphs are used in her work and how it differs from typical graph machine learning use cases. We also discuss her talk from the workshop, “The Why, How, and When of Representations for Complex Systems”, in which Tina argues that one of the reasons practitioners have struggled to model complex systems is because of the lack of connection to the data sourcing and generation process. This is definitely a NERD ALERT approved interview!

The complete show notes for this episode can be found at twimlai.com/go/547

Today we’re excited to kick off our annual NeurIPS, joined by Oriol Vinyals, the lead of the deep learning team at Deepmind. We cover a lot of ground in our conversation with Oriol, beginning with a look at his research agenda and why the scope has remained wide even through the maturity of the field, his thoughts on transformer models and if they will get us beyond the current state of DL, or if some other model architecture would be more advantageous. We also touch on his thoughts on the large language models craze, before jumping into his recent paper StarCraft II Unplugged: Large Scale Offline Reinforcement Learning, a follow up to their popular AlphaStar work from a few years ago. Finally, we discuss the degree to which the work that Deepmind and others are doing around games actually translates into real-world, non-game scenarios, recent work on multimodal few-shot learning, and we close with a discussion of the consequences of the level of scale that we’ve achieved thus far.  

The complete show notes for this episode can be found at twimlai.com/go/546

Today we’re joined by Michael McCourt the head of engineering at SigOpt. In our conversation with Michael, we explore the vast space around the topic of optimization, including the technical differences between ML and optimization and where they’re applied, what the path to increasing complexity looks like for a practitioner and the relationship between optimization and active learning. We also discuss the research frontier for optimization and how folks think about the interesting challenges and open questions for this field, how optimization approaches appeared at the latest NeurIPS conference, and Mike’s excitement for the emergence of interdisciplinary work between the machine learning community and other fields like the natural sciences.

The complete show notes for this episode can be found at twimlai.com/go/545

Today we conclude our AWS re:Invent coverage joined by Brian Granger, a senior principal technologist at Amazon Web Services, and a co-creator of Project Jupyter. In our conversion with Brian, we discuss the inception and early vision of Project Jupyter, including how the explosion of machine learning and deep learning shifted the landscape for the notebook, and how they balanced the needs of these new user bases vs their existing community of scientific computing users. We also explore AWS’s role with Jupyter and why they’ve decided to invest resources in the project, Brian's thoughts on the broader ML tooling space, and how they’ve applied (and the impact of) HCI principles to the building of these tools. Finally, we dig into the recent Sagemaker Canvas and Studio Lab releases and Brian’s perspective on the future of notebooks and the Jupyter community at large.

The complete show notes for this episode can be found at twimlai.com/go/544

Today we continue our 2021 re:Invent series joined by Jack Berkowitz, chief data officer at ADP. In our conversation with Jack, we explore the ever evolving role and growth of machine learning at the company, from the evolution of their ML platform, to the unique team structure. We discuss Jack’s perspective on data governance, the broad use cases for ML, how they approached the decision to move to the cloud, and the impact of scale in the way they deal with data. Finally, we touch on where innovation comes from at ADP, and the challenge of getting the talent it needs to innovate as a large “legacy” company.

The complete show notes for this episode can be found at twimlai.com/go/543

Today we’re joined by Bratin Saha, vice president and general manager at Amazon.

In our conversation with Bratin, we discuss quite a few of the recent ML-focused announcements coming out of last weeks re:Invent conference, including new products like Canvas and Studio Lab, as well as upgrades to existing services like Ground Truth Plus. We explore what no-code environments like the aforementioned Canvas mean for the democratization of ML tooling, and some of the key challenges to delivering it as a consumable product. We also discuss industrialization as a subset of MLOps, and how customer patterns inform the creation of these tools, and much more!

The complete show notes for this episode can be found at twimlai.com/go/542.

Today we’re joined by Doug Burdick, a principal research staff member at IBM Research. In a recent interview, Doug’s colleague Yunyao Li joined us to talk through some of the broader enterprise NLP problems she’s working on. One of those problems is making documents machine consumable, especially with the traditionally archival file type, the PDF. That’s where Doug and his team come in.

In our conversation, we discuss the multimodal approach they’ve taken to identify, interpret, contextualize and extract things like tables from a document, the challenges they’ve faced when dealing with the tables and how they evaluate the performance of models on tables. We also explore how he’s handled generalizing across different formats, how fine-tuning has to be in order to be effective, the problems that appear on the NLP side of things, and how deep learning models are being leveraged within the group.

The complete show notes for this episode can be found at twimlai.com/go/541

Today we’re joined by Shayan Mortazavi, a data science manager at Accenture. 

In our conversation with Shayan, we discuss his talk from the recent SigOpt HPC & AI Summit, titled A Novel Framework Predictive Maintenance Using Dl and Reliability Engineering. In the talk, Shayan proposes a novel deep learning-based approach for prognosis prediction of oil and gas plant equipment in an effort to prevent critical damage or failure. We explore the evolution of reliability engineering, the decision to use a residual-based approach rather than traditional anomaly detection to determine when an anomaly was happening, the challenges of using LSTMs when building these models, the amount of human labeling required to build the models, and much more!

The complete show notes for this episode can be found at twimlai.com/go/540

Today we’re joined by friend-of-the-show Nasrin Mostafazadeh, co-founder of Verneek. 

Though Verneek is still in stealth, Nasrin was gracious enough to share a bit about the company, including their goal of enabling anyone to make data-informed decisions without the need for a technical background, through the use of innovative human-machine interfaces. In our conversation, we explore the state of AI research in the domains relevant to the problem they’re trying to solve and how they use those insights to inform and prioritize their research agenda. We also discuss what advice Nasrin would give to someone thinking about starting a deep tech startup or going from research to product development. 

The complete show notes for today’s show can be found at twimlai.com/go/539.

Today we’re joined by Julie Shah, a professor at the Massachusetts Institute of Technology (MIT). Julie’s work lies at the intersection of aeronautics, astronautics, and robotics, with a specific focus on collaborative and interactive robotics. In our conversation, we explore how robots would achieve the ability to predict what their human collaborators are thinking, what the process of building knowledge into these systems looks like, and her big picture idea of developing a field robot that doesn’t “require a human to be a robot” to work with it. We also discuss work Julie has done on cross-training between humans and robots with the focus on getting them to co-learn how to work together, as well as future projects that she’s excited about.

The complete show notes for this episode can be found at twimlai.com/go/538.

Today we’re joined by Yunyao Li, a senior research manager at IBM Research. 

Yunyao is in a somewhat unique position at IBM, addressing the challenges of enterprise NLP in a traditional research environment, while also having customer engagement responsibilities. In our conversation with Yunyao, we explore the challenges associated with productizing NLP in the enterprise, and if she focuses on solving these problems independent of one another, or through a more unified approach. 

We then ground the conversation with real-world examples of these enterprise challenges, including enabling level document discovery at scale using combinations of techniques like deep neural networks and supervised and/or unsupervised learning, and entity extraction and semantic parsing to identify text. Finally, we talk through data augmentation in the context of NLP, and how we enable the humans in-the-loop to generate high-quality data.

The complete show notes for this episode can be found at twimlai.com/go/537

Today we’re joined by Kim Branson, the SVP and global head of artificial intelligence and machine learning at GSK. 

We cover a lot of ground in our conversation, starting with a breakdown of GSK’s core pharmaceutical business, and how ML/AI fits into that equation, use cases that appear using genetics data as a data source, including sequential learning for drug discovery. We also explore the 500 billion node knowledge graph Kim’s team built to mine scientific literature, and their “AI Hub”, the ML/AI infrastructure team that handles all tooling and engineering problems within their organization. Finally, we explore their recent cancer research collaboration with King’s College, which is tasked with understanding the individualized needs of high- and low-risk cancer patients using ML/AI amongst other technologies. 

The complete show notes for this episode can be found at twimlai.com/go/536.

Today we’re joined by David Ha, a research scientist at Google. 

In nature, there are many examples of “bottlenecks”, or constraints, that have shaped our development as a species. Building upon this idea, David posits that these same evolutionary bottlenecks could work when training neural network models as well. In our conversation with David, we cover a TON of ground, including the aforementioned biological inspiration for his work, then digging deeper into the different types of constraints he’s applied to ML systems. We explore abstract generative models and how advanced training agents inside of generative models has become, and quite a few papers including Neuroevolution of self-interpretable agents, World Models and Attention for Reinforcement Learning, and The Sensory Neuron as a Transformer: Permutation-Invariant Neural Networks for Reinforcement Learning.

This interview is Nerd Alert certified, so get your notes ready! 

PS. David is one of our favorite follows on Twitter (@hardmaru), so check him out and share your thoughts on this interview and his work!

The complete show notes for this episode can be found at twimlai.com/go/535

Today we’re joined by Luke Stark, an assistant professor at Western University in London, Ontario. 

In our conversation with Luke, we explore the existence and use of facial recognition technology, something Luke has been critical of in his work over the past few years, comparing it to plutonium. We discuss Luke’s recent paper, “Physiognomic Artificial Intelligence”, in which he critiques studies that will attempt to use faces and facial expressions and features to make determinations about people, a practice fundamental to facial recognition, also one that Luke believes is inherently racist at its core. 

Finally, briefly discuss the recent wave of hires at the FTC, and the news that broke (mid-recording) announcing that Facebook will be shutting down their facial recognition system and why it's not necessarily the game-changing announcement it seemed on its… face. 

The complete show notes for this episode can be found at twimlai.com/go/534.

Today we’re joined by Francesc Joan Riera, an applied machine learning engineer at The LEGO Group. 

In our conversation, we explore the ML infrastructure at LEGO, specifically around two use cases, content moderation and user engagement. While content moderation is not a new or novel task, but because their apps and products are marketed towards children, their need for heightened levels of moderation makes it very interesting. 

We discuss if the moderation system is built specifically to weed out bad actors or passive behaviors if their system has a human-in-the-loop component, why they built a feature store as opposed to a traditional database, and challenges they faced along that journey. We also talk through the range of skill sets on their team, the use of MLflow for experimentation, the adoption of AWS for serverless, and so much more!

The complete show notes for this episode can be found at twimlai.com/go/534.

Today we’re joined by Hamel Husain, Staff Machine Learning Engineer at GitHub. 

Over the last few years, Hamel has had the opportunity to work on some of the most popular open source projects in the ML world, including fast.ai, nbdev, fastpages, and fastcore, just to name a few. In our conversation with Hamel, we discuss his journey into Silicon Valley, and how he discovered that the ML tooling and infrastructure weren’t quite as advanced as he’d assumed, and how that led him to help build some of the foundational pieces of Airbnb’s Bighead Platform. 

We also spend time exploring Hamel’s time working with Jeremy Howard and the team creating fast.ai, how nbdev came about, and how it plans to change the way practitioners interact with traditional jupyter notebooks. Finally, talk through a few more tools in the fast.ai ecosystem, fastpages, fastcore, how these tools interact with Github Actions, and the up and coming ML tools that Hamel is excited about. 

The complete show notes for this episode can be found at twimlai.com/go/532.

In today’s episode, we are joined by Julianna Ianni, vice president of AI research & development at Proscia.

In our conversation, Julianna shares her and her team’s research focused on developing applications that would help make the life of pathologists easier by enabling tasks to quickly and accurately be diagnosed using deep learning and AI.

We also explore their paper “A Pathology Deep Learning System Capable of Triage of Melanoma Specimens Utilizing Dermatopathologist Consensus as Ground Truth”, while talking through how ML aids pathologists in diagnosing Melanoma by building a multitask classifier to distinguish between low-risk and high-risk cases. Finally, we discussed the challenges involved in designing a model that would help in identifying and classifying Melanoma, the results they’ve achieved, and what the future of this work could look like.

The complete show notes for this episode can be found at twimlai.com/go/531.

Today we’re joined by Akshat Kaul, the head of data science and machine learning at Redfin. We’re all familiar with Redfin, but did you know that redfin.com is the largest real estate brokerage site in the US? In our conversation with Akshat, we discuss the history of ML at Redfin and a few of the key use cases that ML is currently being applied to, including recommendations, price estimates, and their “hot homes” feature. We explore their recent foray into building their own internal platform, which they’ve coined “Redeye”, how they’ve built Redeye to support modeling across the business, and how Akshat thinks about the role of the cloud when building and delivering their platform. Finally, we discuss the impact the pandemic has had on ML at the company, and Akshat’s vision for the future of their platform and machine learning at the company more broadly. 

The complete show notes for this episode can be found at twimlai.com/go/530.

Today we’re joined by Edward Raff, chief scientist and head of the machine learning research group at Booz Allen Hamilton. Edward’s work sits at the intersection of machine learning and cybersecurity, with a particular interest in malware analysis and detection. In our conversation, we look at the evolution of adversarial ML over the last few years before digging into Edward’s recently released paper, Adversarial Transfer Attacks With Unknown Data and Class Overlap. In this paper, Edward and his team explore the use of adversarial transfer attacks and how they’re able to lower their success rate by simulating class disparity. Finally, we talk through quite a few future directions for adversarial attacks, including his interest in graph neural networks.

The complete show notes for this episode can be found at twimlai.com/go/529.

Today we’re joined by Andrea Banino, a research scientist at DeepMind. In our conversation with Andrea, we explore his interest in artificial general intelligence by way of episodic memory, the relationship between memory and intelligence, the challenges of applying memory in the context of neural networks, and how to overcome problems of generalization. 

We also discuss his work on the PonderNet, a neural network that “budgets” its computational investment in solving a problem, according to the inherent complexity of the problem, the impetus and goals of this research, and how PonderNet connects to his memory research. 

The complete show notes for this episode can be found at twimlai.com/go/528.

Take our survey at twimlai.com/survey21!

Today we’re joined by Tim Rocktäschel, a research scientist at Facebook AI Research and an associate professor at University College London (UCL). 

Tim’s work focuses on training RL agents in simulated environments, with the goal of these agents being able to generalize to novel situations. Typically, this is done in environments like OpenAI Gym, MuJuCo, or even using Atari games, but these all come with constraints. In Tim’s approach, he utilizes a game called NetHack, which is much more rich and complex than the aforementioned environments.  

In our conversation with Tim, we explore the ins and outs of using NetHack as a training environment, including how much control a user has when generating each individual game and the challenges he's faced when deploying the agents. We also discuss his work on MiniHack, an environment creation framework and suite of tasks that are based on NetHack, and future directions for this research.

The complete show notes for this episode can be found at twimlai.com/go/527.

In this special episode of the show, we’re excited to bring you our conversation with Prashanth Chandrasekar, CEO of Stack Overflow. This interview was recorded as a part of the annual Prosus AI Marketplace event. 

In our discussion with Prashanth, we explore the impact the pandemic has had on Stack Overflow, how they think about community and enable collaboration in over 100 million monthly users from around the world, and some of the challenges they’ve dealt with when managing a community of this scale. We also examine where Stack Overflow is in their AI journey, use cases illustrating how they’re currently utilizing ML, what their role is in the future of AI-based code generation, what other trends they’ve picked up on over the last few years, and how they’re using those insights to forge the path forward.

The complete show notes for this episode can be found at twimlai.com/go/526.

Today we’re joined by Joseph Soriaga, a senior director of technology at Qualcomm. 

In our conversation with Joseph, we focus on a pair of papers that he and his team will be presenting at Globecom later this year. The first, Neural Augmentation of Kalman Filter with Hypernetwork for Channel Tracking, details the use of deep learning to augment an algorithm to address mismatches in models, allowing for more efficient training and making models more interpretable and predictable. The second paper, WiCluster: Passive Indoor 2D/3D Positioning using WiFi without Precise Labels, explores the use of rf signals to infer what the environment looks like, allowing for estimation of a person’s movement. 

We also discuss the ability for machine learning and AI to help enable 5G and make it more efficient for these applications, as well as the scenarios that ML would allow for more effective delivery of connected services, and look towards what might be possible in the near future. 

The complete show notes for this episode can be found at twimlai.com/go/525.

Today we’re joined by Kanaka Rajan, an assistant professor at the Icahn School of Medicine at Mt Sinai. Kanaka, who is a recent recipient of the NSF Career Award, bridges the gap between the worlds of biology and artificial intelligence with her work in computer science. In our conversation, we explore how she builds “lego models” of the brain that mimic biological brain functions, then reverse engineers those models to answer the question “do these follow the same operating principles that the biological brain uses?”

We also discuss the relationship between memory and dynamically evolving system states, how close we are to understanding how memory actually works, how she uses RNNs for modeling these processes, and what training and data collection looks like. Finally, we touch on her use of curriculum learning (where the task you want a system to learn increases in complexity slowly), and of course, we look ahead at future directions for Kanaka’s research. 

The complete show notes for this episode can be found at twimlai.com/go/524.

Today we’re joined by a friend of the show and return guest Ville Tuulos, CEO and co-founder of Outerbounds. In our previous conversations with Ville, we explored his experience building and deploying the open-source framework, Metaflow, while working at Netflix. Since our last chat, Ville has embarked on a few new journeys, including writing the upcoming book Effective Data Science Infrastructure, and commercializing Metaflow, both of which we dig into quite a bit in this conversation. 

We reintroduce the problem that Metaflow was built to solve and discuss some of the unique use cases that Ville has seen since it's release, the relationship between Metaflow and Kubernetes, and the maturity of services like batch and lambdas allowing a complete production ML system to be delivered. Finally, we discuss the degree to which Ville is catering is Outerbounds’ efforts to building tools for the MLOps community, and what the future looks like for him and Metaflow. 

The complete show notes for this episode can be found at twimlai.com/go/523.

Today we’re joined by Li Jiang, a distinguished engineer at Microsoft working on Azure Speech. 

In our conversation with Li, we discuss his journey across 27 years at Microsoft, where he’s worked on, among other things, audio and speech recognition technologies. We explore his thoughts on the advancements in speech recognition over the past few years, the challenges, and advantages, of using either end-to-end or hybrid models. 

We also discuss the trade-offs between delivering accuracy or quality and the kind of runtime characteristics that you require as a service provider, in the context of engineering and delivering a service at the scale of Azure Speech. Finally, we walk through the data collection process for customizing a voice for TTS, what languages are currently supported, managing the responsibilities of threats like deep fakes, the future for services like these, and much more!

The complete show notes for this episode can be found at twimlai.com/go/522.

Today we’re joined by Sandra Wacther, an associate professor and senior research fellow at the University of Oxford. 

Sandra’s work lies at the intersection of law and AI, focused on what she likes to call “algorithmic accountability”. In our conversation, we explore algorithmic accountability in three segments, explainability/transparency, data protection, and bias, fairness and discrimination. We discuss how the thinking around black boxes changes when discussing applying regulation and law, as well as a breakdown of counterfactual explanations and how they’re created. We also explore why factors like the lack of oversight lead to poor self-regulation, and the conditional demographic disparity test that she helped develop to test bias in models, which was recently adopted by Amazon.

The complete show notes for this episode can be found at twimlai.com/go/521.

Today we’re joined by Dillon Erb, CEO of Paperspace. 

If you’re not familiar with Dillon, he joined us about a year ago to discuss Machine Learning as a Software Engineering Discipline; we strongly encourage you to check out that interview as well. In our conversation, we explore the idea of compositional AI, and if it is the next frontier in a string of recent game-changing machine learning developments. We also discuss a source of constant back and forth in the community around the role of notebooks, and why Paperspace made the choice to pivot towards a more traditional engineering code artifact model after building a popular notebook service. Finally, we talk through their newest release Workflows, an automation and build system for ML applications, which Dillon calls their “most ambitious and comprehensive project yet.”

The complete show notes for this episode can be found at twimlai.com/go/520.

Today we’re joined by Yanshuai Cao, a senior research team lead at Borealis AI. In our conversation with Yanshuai, we explore his work on Turing, their natural language to SQL engine that allows users to get insights from relational databases without having to write code. We do a bit of compare and contrast with the recently released Codex Model from OpenAI, the role that reasoning plays in solving this problem, and how it is implemented in the model. We also talk through various challenges like data augmentation, the complexity of the queries that Turing can produce, and a paper that explores the explainability of this model.

The complete show notes for this episode can be found at twimlai.com/go/519.

Today we’re joined by Yejin Choi, a professor at the University of Washington. We had the pleasure of catching up with Yejin after her keynote interview at the recent Stanford HAI “Foundational Models” workshop. In our conversation, we explore her work at the intersection of natural language generation and common sense reasoning, including how she defines common sense, and what the current state of the world is for that research. We discuss how this could be used for creative storytelling, how transformers could be applied to these tasks, and we dig into the subfields of physical and social common sense reasoning. Finally, we talk through the future of Yejin’s research and the areas that she sees as most promising going forward. 

If you enjoyed this episode, check out our conversation on AI Storytelling Systems with Mark Riedl. The complete show notes for today’s episode can be found at twimlai.com/go/518.

Today we’re joined by Konrad Tollmar, research director at Electronic Arts and an associate professor at KTH. 

In our conversation, we explore his role as the lead of EA’s applied research team SEED and the ways that they’re applying ML/AI across popular franchises like Apex Legends, Madden, and FIFA. We break down a few papers focused on the application of ML to game testing, discussing why deep reinforcement learning is at the top of their research agenda, the differences between training atari games and modern 3D games, using CNNs to detect glitches in games, and of course, Konrad gives us his outlook on the future of ML for games training.

The complete show notes for this episode can be found at twimlai.com/go/517.

Today we’re joined by Kai-Fu Lee, chairman and CEO of Sinovation Ventures and author of AI 2041: Ten Visions for Our Future. 

In AI 2041, Kai-Fu and co-author Chen Qiufan tell the story of how AI could shape our future through a series of 10 “scientific fiction” short stories. In our conversation with Kai-Fu, we explore why he chose 20 years as the time horizon for these stories, and dig into a few of the stories in more detail. We explore the potential for level 5 autonomous driving and what effect that will have on both established and developing nations, the potential outcomes when dealing with job displacement, and his perspective on how the book will be received. We also discuss the potential consequences of autonomous weapons, if we should actually worry about singularity or superintelligence, and the evolution of regulations around AI in 20 years.

We’d love to hear from you! What are your thoughts on any of the stories we discuss in the interview? Will you be checking this book out? Let us know in the comments on the show notes page at twimlai.com/go/516.

Today we’re joined by Daniela Rus, director of CSAIL & Deputy Dean of Research at MIT. 

In our conversation with Daniela, we explore the history of CSAIL, her role as director of one of the most prestigious computer science labs in the world, how she defines robots, and her take on the current AI for robotics landscape. We also discuss some of her recent research interests including soft robotics, adaptive control in autonomous vehicles, and a mini surgeon robot made with sausage casing(?!). 

The complete show notes for this episode can be found at twimlai.com/go/515.

Today we’re joined by Alexander Richard, a research scientist at Facebook Reality Labs, and recipient of the ICLR Best Paper Award for his paper “Neural Synthesis of Binaural Speech From Mono Audio.” 

We begin our conversation with a look into the charter of Facebook Reality Labs, and Alex’s specific Codec Avatar project, where they’re developing AR/VR for social telepresence (applications like this come to mind). Of course, we dig into the aforementioned paper, discussing the difficulty in improving the quality of audio and the role of dynamic time warping, as well as the challenges of creating this model. Finally, Alex shares his thoughts on 3D rendering for audio, and other future research directions. 

The complete show notes for this episode can be found at twimlai.com/go/514.

Today we’re joined by Alona Fyshe, an assistant professor at the University of Alberta. 

We caught up with Alona on the heels of an interesting panel discussion that she participated in, centered around improving AI systems using research about brain activity. In our conversation, we explore the multiple types of brain images that are used in this research, what representations look like in these images, and how we can improve language models without knowing explicitly how the brain understands the language. We also discuss similar experiments that have incorporated vision, the relationship between computer vision models and the representations that language models create, and future projects like applying a reinforcement learning framework to improve language generation.

The complete show notes for this episode can be found at twimlai.com/go/513.

Today we’re joined by Samory Kpotufe, an associate professor at Columbia University and program chair of the 2021 Conference on Learning Theory (COLT). 

In our conversation with Samory, we explore his research at the intersection of machine learning, statistics, and learning theory, and his goal of reaching self-tuning, adaptive algorithms. We discuss Samory’s research in transfer learning and other potential procedures that could positively affect transfer, as well as his work understanding unsupervised learning including how clustering could be applied to real-world applications like cybersecurity, IoT (Smart homes, smart city sensors, etc) using methods like dimension reduction, random projection, and others. If you enjoyed this interview, you should definitely check out our conversation with Jelani Nelson on the “Theory of Computation.” 

The complete show notes for this episode can be found at https://twimlai.com/go/512.

Today we’re joined by Eric Rice, associate professor at USC, and the co-director of the USC Center for Artificial Intelligence in Society. 

Eric is a sociologist by trade, and in our conversation, we explore how he has made extensive inroads within the machine learning community through collaborations with ML academics and researchers. We discuss some of the most important lessons Eric has learned while doing interdisciplinary projects, how the social scientist’s approach to assessment and measurement would be different from a computer scientist's approach to assessing the algorithmic performance of a model. 

We specifically explore a few projects he’s worked on including HIV prevention amongst the homeless youth population in LA, a project he spearheaded with former guest Milind Tambe, as well as a project focused on using ML techniques to assist in the identification of people in need of housing resources, and ensuring that they get the best interventions possible. 

If you enjoyed this conversation, I encourage you to check out our conversation with Milind Tambe from last year’s TWIMLfest on Why AI Innovation and Social Impact Go Hand in Hand.

The complete show notes for this episode can be found at https://twimlai.com/go/511.

Today we’re joined by José Miguel Hernández-Lobato, a university lecturer in machine learning at the University of Cambridge. In our conversation with Miguel, we explore his work at the intersection of Bayesian learning and deep learning. We discuss how he’s been applying this to the field of molecular design and discovery via two different methods, with one paper searching for possible chemical reactions, and the other doing the same, but in 3D and in 3D space. We also discuss the challenges of sample efficiency, creating objective functions, and how those manifest themselves in these experiments, and how he integrated the Bayesian approach to RL problems. We also talk through a handful of other papers that Miguel has presented at recent conferences, which are all linked at twimlai.com/go/510.

Today we’re joined by return guest Greg Brockman, co-founder and CTO of OpenAI. We had the pleasure of reconnecting with Greg on the heels of the announcement of Codex, OpenAI’s most recent release. Codex is a direct descendant of GPT-3 that allows users to do autocomplete tasks based on all of the publicly available text and code on the internet. In our conversation with Greg, we explore the distinct results Codex sees in comparison to GPT-3, relative to the prompts it's being given, how it could evolve given different types of training data, and how users and practitioners should think about interacting with the API to get the most out of it. We also discuss Copilot, their recent collaboration with Github that is built on Codex, as well as the implications of Codex on coding education, explainability, and broader societal issues like fairness and bias, copyrighting, and jobs. 

The complete show notes for this episode can be found at twimlai.com/go/509.

Today we’re joined by Rose Yu, an assistant professor at the Jacobs School of Engineering at UC San Diego. 

Rose’s research focuses on advancing machine learning algorithms and methods for analyzing large-scale time-series and spatial-temporal data, then applying those developments to climate, transportation, and other physical sciences. We discuss how Rose incorporates physical knowledge and partial differential equations in these use cases and how symmetries are being exploited. We also explore their novel neural network design that is focused on non-traditional convolution operators and allows for general symmetry, how we get from these representations to the network architectures that she has developed and another recent paper on deep spatio-temporal models. 

The complete show note for this episode can be found at twimlai.com/go/508.

Today we’re joined by Bryan Catanzaro, vice president of applied deep learning research at NVIDIA.

Most folks know Bryan as one of the founders/creators of cuDNN, the accelerated library for deep neural networks. In our conversation, we explore his interest in high-performance computing and its recent overlap with AI, his current work on Megatron, a framework for training giant language models, and the basic approach for distributing a large language model on DGX infrastructure. 

We also discuss the three different kinds of parallelism, tensor parallelism, pipeline parallelism, and data parallelism, that Megatron provides when training models, as well as his work on the Deep Learning Super Sampling project and the role it's playing in the present and future of game development via ray tracing. 

The complete show notes for this episode can be found at twimlai.com/go/507.

Today we close out our 2021 ICML series joined by Lina Montoya, a postdoctoral researcher at UNC Chapel Hill. 

In our conversation with Lina, who was an invited speaker at the Neglected Assumptions in Causal Inference Workshop, we explored her work applying Optimal Dynamic Treatment (ODT) to understand which kinds of individuals respond best to specific interventions in the US criminal justice system. We discuss the concept of neglected assumptions and how it connects to ODT rule estimation, as well as a breakdown of the causal roadmap, coined by researchers at UC Berkeley. 

Finally, Lina talks us through the roadmap while applying the ODT rule problem, how she’s applied a “superlearner” algorithm to this problem, how it was trained, and what the future of this research looks like.

The complete show notes for this episode can be found at twimlai.com/go/506.

Today we continue our ICML series joined by Gustavo Malkomes, a research engineer at Intel via their recent acquisition of SigOpt. 

In our conversation with Gustavo, we explore his paper Beyond the Pareto Efficient Frontier: Constraint Active Search for Multiobjective Experimental Design, which focuses on a novel algorithmic solution for the iterative model search process. This new algorithm empowers teams to run experiments where they are not optimizing particular metrics but instead identifying parameter configurations that satisfy constraints in the metric space. This allows users to efficiently explore multiple metrics at once in an efficient, informed, and intelligent way that lends itself to real-world, human-in-the-loop scenarios.

The complete show notes for this episode can be found at twimlai.com/go/505.

Today we kick off our ICML coverage joined by Virginia Smith, an assistant professor in the Machine Learning Department at Carnegie Mellon University. 

In our conversation with Virginia, we explore her work on cross-device federated learning applications, including where the distributed learning aspects of FL are relative to the privacy techniques. We dig into her paper from ICML, Ditto: Fair and Robust Federated Learning Through Personalization, what fairness means in contrast to AI ethics, the particulars of the failure modes, the relationship between models, and the things being optimized across devices, and the tradeoffs between fairness and robustness.

We also discuss a second paper, Heterogeneity for the Win: One-Shot Federated Clustering, how the proposed method makes heterogeneity beneficial in data, how the heterogeneity of data is classified, and some applications of FL in an unsupervised setting.

The complete show notes for this episode can be found at twimlai.com/go/504.

Today we’re joined by Errol Koolmeister, the head of AI foundation at H&M Group.

In our conversation with Errol, we explore H&M’s AI journey, including its wide adoption across the company in 2016, and the various use cases in which it's deployed like fashion forecasting and pricing algorithms. We discuss Errol’s first steps in taking on the challenge of scaling AI broadly at the company, the value-added learning from proof of concepts, and how to align in a sustainable, long-term way. Of course, we dig into the infrastructure and models being used, the biggest challenges faced, and the importance of managing the project portfolio, while Errol shares their approach to building infra for a specific product with many products in mind.

Today we’re joined by Stefano Soatto, VP of AI applications science at AWS and a professor of computer science at UCLA. 

Our conversation with Stefano centers on recent research of his called Graceful AI, which focuses on how to make trained systems evolve gracefully. We discuss the broader motivation for this research and the potential dangers or negative effects of constantly retraining ML models in production. We also talk about research into error rate clustering, the importance of model architecture when dealing with problems of model compression, how they’ve solved problems of regression and reprocessing by utilizing existing models, and much more.

The complete show notes for this episode can be found at twimlai.com/go/502.

Today we’re joined by Suchi Saria, the founder and CEO of Bayesian Health, the John C. Malone associate professor of computer science, statistics, and health policy, and the director of the machine learning and healthcare lab at Johns Hopkins University. 

Suchi shares a bit about her journey to working in the intersection of machine learning and healthcare, and how her research has spanned across both medical policy and discovery. We discuss why it has taken so long for machine learning to become accepted and adopted by the healthcare infrastructure and where exactly we stand in the adoption process, where there have been “pockets” of tangible success. 

Finally, we explore the state of healthcare data, and of course, we talk about Suchi’s recently announced startup Bayesian Health and their goals in the healthcare space, and an accompanying study that looks at real-time ML inference in an EMR setting.

The complete show notes for this episode can be found at twimlai.com/go/501.

Today we’re joined by a friend of the show Jeff Gehlhaar, VP of technology and the head of AI software platforms at Qualcomm. 

In our conversation with Jeff, we cover a ton of ground, starting with a bit of exploration around ML compilers, what they are, and their role in solving issues of parallelism. We also dig into the latest additions to the Snapdragon platform, AI Engine Direct, and how it works as a bridge to bring more capabilities across their platform, how benchmarking works in the context of the platform, how the work of other researchers we’ve spoken to on compression and quantization finds its way from research to product, and much more! 

After you check out this interview, you can look below for some of the other conversations with researchers mentioned. 

The complete show notes for this episode can be found at twimlai.com/go/500.

Today we continue our AI in Innovation series joined by Dan Bohus, senior principal researcher at Microsoft Research, and Siddhartha Sen, a principal researcher at Microsoft Research. 

In this conversation, we use a pair of research projects, Maia Chess and Situated Interaction, to springboard us into a conversation about the evolution of human-AI interaction. We discuss both of these projects individually, as well as the commonalities they have, how themes like understanding the human experience appear in their work, the types of models being used, the various types of data, and the complexity of each of their setups. 

We explore some of the challenges associated with getting computers to better understand human behavior and interact in ways that are more fluid. Finally, we touch on what excites both Dan and Sid about their respective projects, and what they’re excited about for the future.  

The complete show notes for this episode can be found at https://twimlai.com/go/499.

Today we’re joined by Julieta Martinez, a senior research scientist at recently announced startup Waabi. 

Julieta was a keynote speaker at the recent LatinX in AI workshop at CVPR, and our conversation focuses on her talk “What do Large-Scale Visual Search and Neural Network Compression have in Common,” which shows that multiple ideas from large-scale visual search can be used to achieve state-of-the-art neural network compression. We explore the commonality between large databases and dealing with high dimensional, many-parameter neural networks, the advantages of using product quantization, and how that plays out when using it to compress a neural network. 

We also dig into another paper Julieta presented at the conference, Deep Multi-Task Learning for Joint Localization, Perception, and Prediction, which details an architecture that is able to reuse computation between the three tasks, and is thus able to correct localization errors efficiently.

The complete show notes for this episode can be found at twimlai.com/go/498.

Today we continue our CVPR 2021 coverage joined by Claire Monteleoni, an associate professor at the University of Colorado Boulder. 

We cover quite a bit of ground in our conversation with Claire, including her journey down the path from environmental activist to one of the leading climate informatics researchers in the world. We explore her current research interests, and the available opportunities in applying machine learning to climate informatics, including the interesting position of doing ML from a data-rich environment. 

Finally, we dig into the evolution of climate science-focused events and conferences, as well as the Keynote Claire gave at the EarthVision workshop at CVPR “Deep Unsupervised Learning for Climate Informatics,” which focused on semi- and unsupervised deep learning approaches to studying rare and extreme climate events.

The complete show notes for this episode can be found at twimlai.com/go/497.

Today we kick off our CVPR coverage joined by Amir Habibian, a senior staff engineer manager at Qualcomm Technologies. 

In our conversation with Amir, whose research primarily focuses on video perception, we discuss a few papers they presented at the event. We explore the paper Skip-Convolutions for Efficient Video Processing, which looks at training discrete variables to end to end into visual neural networks. We also discuss his work on his FrameExit paper, which proposes a conditional early exiting framework for efficient video recognition. 

The complete show notes for this episode can be found at twimlai.com/go/496.

Today we’re joined by Noam Slonim, the principal investigator of Project Debater at IBM Research. 

In our conversation with Noam, we explore the history of Project Debater, the first AI system that can “debate” humans on complex topics. We also dig into the evolution of the project, which is the culmination of 7 years and over 50 research papers, and eventually becoming a Nature cover paper, “An Autonomous Debating System,” which details the system in its entirety. 

Finally, Noam details many of the underlying capabilities of Debater, including the relationship between systems preparation and training, evidence detection, detecting the quality of arguments, narrative generation, the use of conventional NLP methods like entity linking, and much more.

The complete show notes for this episode can be found at twimlai.com/go/495.

Today we’re joined by Madhur Behl, an Assistant Professor in the department of computer science at the University of Virginia. 

In our conversation with Madhur, we explore the super interesting work he’s doing at the intersection of autonomous driving, ML/AI, and Motorsports, where he’s teaching self-driving cars how to drive in an agile manner. We talk through the differences between traditional self-driving problems and those encountered in a racing environment, the challenges in solving planning, perception, control. 

We also discuss their upcoming race at the Indianapolis Motor Speedway, where Madhur and his students will compete for 1 million dollars in the world’s first head-to-head fully autonomous race, and how they’re preparing for it.

Today we continue our AI Innovation series joined by Microsoft’s Chief Scientific Officer, Eric Horvitz. 

In our conversation with Eric, we explore his tenure as AAAI president and his focus on the future of AI and its ethical implications, the scope of the study on the topic, and how drastically the AI and machine learning landscape has changed since 2009. We also discuss Eric’s role at Microsoft and the Aether committee that has advised the company on issues of responsible AI since 2017.

Finally, we talk through his recent work as a member of the National Security Commission on AI, where he helped commission a 750+ page report on topics including the Future of AI R&D, Building Trustworthy AI systems, civil liberties and privacy, and the challenging area of AI and autonomous weapons.  

The complete show notes for this episode can be found at twimlai.com/go/493.

Today we’re joined by Parvez Ahammad, head of data science applied research at LinkedIn.

In our conversation, Parvez shares his interesting take on organizing principles for his organization, starting with how data science teams are broadly organized at LinkedIn. We explore how they ensure time investments on long-term projects are managed, how to identify products that can help in a cross-cutting way across multiple lines of business, quantitative methodologies to identify unintended consequences in experimentation, and navigating the tension between research and applied ML teams in an organization. Finally, we discuss differential privacy, and their recently released GreyKite library, an open-source Python library developed to support forecasting.

The complete show note for this episode can be found at twimlai.com/go/492.

Today we’re joined Katherine J. Kuchenbecker, director at the Max Planck Institute for Intelligent Systems and of the haptic intelligence department. 

In our conversation, we explore Katherine’s research interests, which lie at the intersection of haptics (physical interaction with the world) and machine learning, introducing us to the concept of “haptic intelligence.” We discuss how ML, mainly computer vision, has been integrated to work together with robots, and some of the devices that Katherine’s lab is developing to take advantage of this research.

We also talk about hugging robots, augmented reality in robotic surgery, and the degree to which she studies human-robot interaction. Finally, Katherine shares with us her passion for mentoring and the importance of diversity and inclusion in robotics and machine learning. 

The complete show notes for this episode can be found at twimlai.com/go/491.

Today we continue our coverage of the AWS ML Summit joined by Chris Fregly, a principal developer advocate at AWS, and Antje Barth, a senior developer advocate at AWS. 

In our conversation with Chris and Antje, we explore their roles as community builders prior to, and since, joining AWS, as well as their recently released book Data Science on AWS. In the book, Chris and Antje demonstrate how to reduce cost and improve performance while successfully building and deploying data science projects. 

We also discuss the release of their new Practical Data Science Specialization on Coursera, managing the complexity that comes with building real-world projects, and some of their favorite sessions from the recent ML Summit.

Today we’re joined by Ziad Asghar, vice president of product management for snapdragon technologies & roadmap at Qualcomm Technologies. 

We begin our conversation with Ziad exploring the symbiosis between 5G and AI and what is enabling developers to take full advantage of AI on mobile devices. We also discuss the balance of product evolution and incorporating research concepts, and the evolution of their hardware infrastructure Cloud AI 100, their role in the deployment of Ingenuity, the robotic helicopter that operated on Mars just last year. 

Finally, we talk about specialization in building IoT applications like autonomous vehicles and smart cities, the degree to which federated learning is being deployed across the industry, and the importance of privacy and security of personal data. 

The complete show notes can be found at https://twimlai.com/go/489.

Today we’re joined by Nir Bar-Lev, co-founder and CEO of ClearML.

In our conversation with Nir, we explore how his view of the wide vs deep machine learning platforms paradox has changed and evolved over time, how companies should think about building vs buying and integration, and his thoughts on why experiment management has become an automatic buy, be it open source or otherwise. 

We also discuss the disadvantages of using a cloud vendor as opposed to a software-based approach, the balance between mlops and data science when addressing issues of overfitting, and how ClearML is applying techniques like federated machine learning and transfer learning to their solutions.

The complete show notes for this episode can be found at https://twimlai.com/go/488.

Today we’re joined by Alex Smola, Vice President and Distinguished Scientist at AWS AI.

We had the pleasure to catch up with Alex prior to the upcoming AWS Machine Learning Summit, and we covered a TON of ground in the conversation. We start by focusing on his research in the domain of deep learning on graphs, including a few examples showcasing its function, and an interesting discussion around the relationship between large language models and graphs. Next up, we discuss their focus on AutoML research and how it's the key to lowering the barrier of entry for machine learning research.

Alex also shares a bit about his work on causality and causal modeling, introducing us to the concept of Granger causality. Finally, we talk about the aforementioned ML Summit, its exponential growth since its inception a few years ago, and what speakers he's most excited about hearing from.

The complete show notes for this episode can be found at https://twimlai.com/go/487.

Today we’re joined by Sean Taylor, Staff Data Scientist at Lyft Rideshare Labs.

We cover a lot of ground with Sean, starting with his recent decision to step away from his previous role as the lab director to take a more hands-on role, and what inspired that change. We also discuss his research at Rideshare Labs, where they take a more “moonshot” approach to solving the typical problems like forecasting and planning, marketplace experimentation, and decision making, and how his statistical approach manifests itself in his work.

Finally, we spend quite a bit of time exploring the role of causality in the work at rideshare labs, including how systems like the aforementioned forecasting system are designed around causal models, if driving model development is more effective using business metrics, challenges associated with hierarchical modeling, and much much more.

The complete show notes for this episode can be found at twimlai.com/go/486.

Today we’re joined by Jabran Zahid, a Senior Researcher at Microsoft Research.

In our conversation with Jabran, we explore their recent endeavor into the complete mapping of which T-cells bind to which antigens through the Antigen Map Project. We discuss how Jabran’s background in astrophysics and cosmology has translated to his current work in immunology and biology, the origins of the antigen map, the biological and how the focus was changed by the emergence of the coronavirus pandemic.

We talk through the biological advancements, and the challenges of using machine learning in this setting, some of the more advanced ML techniques that they’ve tried that have not panned out (as of yet), the path forward for the antigen map to make a broader impact, and much more.

The complete show notes for this episode can be found at twimlai.com/go/485.

Today we conclude our 2021 ICLR coverage joined by Konstantin Rusch, a PhD Student at ETH Zurich.

In our conversation with Konstantin, we explore his recent papers, titled coRNN and uniCORNN respectively, which focus on a novel architecture of recurrent neural networks for learning long-time dependencies.

We explore the inspiration he drew from neuroscience when tackling this problem, how the performance results compared to networks like LSTMs and others that have been proven to work on this problem and Konstantin’s future research goals.

The complete show notes for this episode can be found at twimlai.com/go/484.

Today we continue our ICLR ‘21 series joined by Allyson Ettinger, an Assistant Professor at the University of Chicago. 

One of our favorite recurring conversations on the podcast is the two-way street that lies between machine learning and neuroscience, which Allyson explores through the modeling of cognitive processes that pertain to language. In our conversation, we discuss how she approaches assessing the competencies of AI, the value of control of confounding variables in AI research, and how the pattern matching traits of Ml/DL models are not necessarily exclusive to these systems. 

Allyson also participated in a recent panel discussion at the ICLR workshop How Can Findings About The Brain Improve AI Systems?, centered around the utility of brain inspiration for developing AI models. We discuss ways in which we can try to more closely simulate the functioning of a brain, where her work fits into the analysis and interpretability area of NLP, and much more!

The complete show notes for this episode can be found at twimlai.com/go/483. 

Today we kick off our ICLR 2021 coverage joined by Roberto Bondesan, an AI Researcher at Qualcomm. 

In our conversation with Roberto, we explore his paper Probabilistic Numeric Convolutional Neural Networks, which represents features as Gaussian processes, providing a probabilistic description of discretization error. We discuss some of the other work the team at Qualcomm presented at the conference, including a paper called Adaptive Neural Compression, as well as work on Guage Equvariant Mesh CNNs. Finally, we briefly discuss quantum deep learning, and what excites Roberto and his team about the future of their research in combinatorial optimization.  

The complete show notes for this episode can be found at https://twimlai.com/go/482

Today we’re joined by Huiji Gao, a Senior Engineering Manager of Machine Learning and AI at LinkedIn. 

In our conversation with Huiji, we dig into his interest in building NLP tools and systems, including a recent open-source project called DeText, a framework for generating models for ranking classification and language generation. We explore the motivation behind DeText, the landscape at LinkedIn before and after it was put into use broadly, and the various contexts it’s being used in at the company. We also discuss the relationship between BERT and DeText via LiBERT, a version of BERT that is trained and calibrated on LinkedIn data, the practical use of these tools from an engineering perspective, the approach they’ve taken to optimization, and much more!

The complete show notes for this episode can be found at https://twimlai.com/go/481. 

Today we’re joined by Jacqueline Nolis, Head of Data Science at Saturn Cloud, and co-host of the Build a Career in Data Science Podcast. 

You might remember Jacqueline from our Advancing Your Data Science Career During the Pandemic panel, where she shared her experience trying to navigate the suddenly hectic data science job market. Now, a year removed from that panel, we explore her book on data science careers, top insights for folks just getting into the field, ways that job seekers should be signaling that they have the required background, and how to approach and navigate failure as a data scientist. 

We also spend quite a bit of time discussing Dask, an open-source library for parallel computing in Python, as well as use cases for the tool, the relationship between dask and Kubernetes and docker containers, where data scientists are in regards to the software development toolchain and much more!

The complete show notes for this episode can be found at https://twimlai.com/go/480.

Today we’re joined by Irene Chen, a Ph.D. student at MIT. 

Irene’s research is focused on developing new machine learning methods specifically for healthcare, through the lens of questions of equity and inclusion. In our conversation, we explore some of the various projects that Irene has worked on, including an early detection program for intimate partner violence. 

We also discuss how she thinks about the long term implications of predictions in the healthcare domain, how she’s learned to communicate across the interface between the ML researcher and clinician, probabilistic approaches to machine learning for healthcare, and finally, key takeaways for those of you interested in this area of research.

The complete show notes for this episode can be found at https://twimlai.com/go/479.

Today we’re joined by Mark Riedl, a Professor in the School of Interactive Computing at Georgia Tech. In our conversation with Mark, we explore his work building AI storytelling systems, mainly those that try and predict what listeners think will happen next in a story and how he brings together many different threads of ML/AI together to solve these problems. We discuss how the theory of mind is layered into his research, the use of large language models like GPT-3, and his push towards being able to generate suspenseful stories with these systems. 

We also discuss the concept of intentional creativity and the lack of good theory on the subject, the adjacent areas in ML that he’s most excited about for their potential contribution to his research, his recent focus on model explainability, how he approaches problems of common sense, and much more! 

The complete show notes for this episode can be found at https://twimlai.com/go/478.

Today we’re joined by Jamie Macbeth, an assistant professor in the department of computer science at Smith College. 

In our conversation with Jamie, we explore his work at the intersection of cognitive systems and natural language understanding, and how to use AI as a vehicle for better understanding human intelligence. We discuss the tie that binds these domains together, if the tasks are the same as traditional NLU tasks, and what are the specific things he’s trying to gain deeper insights into.

One of the unique aspects of Jamie’s research is that he takes an “old-school AI” approach, and to that end, we discuss the models he handcrafts to generate language. Finally, we examine how he evaluates the performance of his representations if he’s not playing the SOTA “game,” what he bookmarks against, identifying deficiencies in deep learning systems, and the exciting directions for his upcoming research. 

The complete show notes for this episode can be found at https://twimlai.com/go/477.

Today we’re joined by Pieter Abbeel, a Professor at UC Berkeley, co-Director of the Berkeley AI Research Lab (BAIR), as well as Co-founder and Chief Scientist at Covariant.

In our conversation with Pieter, we cover a ton of ground, starting with the specific goals and tasks of his work at Covariant, the shift in needs for industrial AI application and robots, if his experience solving real-world problems has changed his opinion on end to end deep learning, and the scope for the three problem domains of the models he’s building.

We also explore his recent work at the intersection of unsupervised and reinforcement learning, goal-directed RL, his recent paper “Pretrained Transformers as Universal Computation Engines” and where that research thread is headed, and of course, his new podcast Robot Brains, which you can find on all streaming platforms today!

The complete show notes for this episode can be found at twimlai.com/go/476.

Today we’re joined by Abhishek Thakur, a machine learning engineer at Hugging Face, and the world’s first Quadruple Kaggle Grandmaster!

In our conversation with Abhishek, we explore his Kaggle journey, including how his approach to competitions has evolved over time, what resources he used to prepare for his transition to a full-time practitioner, and the most important lessons he’s learned along the way.

We also spend a great deal of time discussing his new role at HuggingFace, where he's building AutoNLP. We talk through the goals of the project, the primary problem domain, and how the results of AutoNLP compare with those from hand-crafted models. Finally, we discuss Abhishek’s book, Approaching (Almost) Any Machine Learning Problem.

The complete show notes for this episode can be found at https://twimlai.com/go/475.

Today we’re joined by Saqib Shaikh, a Software Engineer at Microsoft, and the lead for the Seeing AI Project.

In our conversation with Saqib, we explore the Seeing AI app, an app “that narrates the world around you.” We discuss the various technologies and use cases for the app, and how it has evolved since the inception of the project, how the technology landscape supports projects like this one, and the technical challenges he faces when building out the app.

We also the relationship and trust between humans and robots, and how that translates to this app, what Saqib sees on the research horizon that will support his vision for the future of Seeing AI, and how the integration of tech like Apple’s upcoming “smart” glasses could change the way their app is used.

The complete show notes for this episode can be found at twimlai.com/go/474.

Today we’re joined by Jelani Nelson, a professor in the Theory Group at UC Berkeley.

In our conversation with Jelani, we explore his research in computational theory, where he focuses on building streaming and sketching algorithms, random projections, and dimensionality reduction. We discuss how Jelani thinks about the balance between the innovation of new algorithms and the performance of existing ones, and some use cases where we’d see his work in action.

Finally, we talk through how his work ties into machine learning, what tools from the theorist’s toolbox he’d suggest all ML practitioners know, and his nonprofit AddisCoder, a 4 week summer program that introduces high-school students to programming and algorithms.

The complete show notes for this episode can be found at twimlai.com/go/473.

Today we’re joined by Stevie Chancellor, an Assistant Professor in the Department of Computer Science and Engineering at the University of Minnesota.

In our conversation with Stevie, we explore her work at the intersection of human-centered computing, machine learning, and high-risk mental illness behaviors. We discuss how her background in HCC helps shapes her perspective, how machine learning helps with understanding severity levels of mental illness, and some recent work where convolutional graph neural networks are applied to identify and discover new kinds of behaviors for people who struggle with opioid use disorder.

We also explore the role of computational linguistics and NLP in her research, issues in using social media data being used as a data source, and finally, how people who are interested in an introduction to human-centered computing can get started.

The complete show notes for this episode can be found at twimlai.com/go/472.

In this episode, we’re joined by Dataiku’s Director of Data Science, Conor Jensen. In our conversation, we explore the panel he lead at TWIMLcon “AI Operationalization: Where the AI Rubber Hits the Road for the Enterprise,” discussing the ML journey of each panelist’s company, and where Dataiku fits in the equation.

The complete show notes for this episode can be found at https://twimlai.com/go/471. 

In this episode, we’re joined by Diego Oppenheimer, Founder and CEO of Algorithmia. In our conversation, we discuss Algorithmia’s involvement with TWIMLcon, as well as an exploration of the results of their recently conducted survey on the state of the AI market.

The complete show notes for this episode can be found at twimlai.com/go/470.

In this episode, we’re joined by Santiago Giraldo, Director Of Product Marketing for Data Engineering & Machine Learning at Cloudera. In our conversation, we discuss Cloudera’s talks at TWIMLcon, as well as their various research efforts from their Fast Forward Labs arm.

The complete show notes for this episode can be found at twimlai.com/sponsorseries.

In this episode, we’re joined by Paul van der Boor, Senior Director of Data Science at Prosus, to discuss his TWIMLcon experience and how they’re using ML platforms to manage machine learning at a global scale.

The complete show notes for this episode can be found at twimlai.com/sponsorseries.

Today we’re joined by Emily M. Bender, Professor at the University of Washington, and AI Researcher, Margaret Mitchell. 

Emily and Meg, as well as Timnit Gebru and Angelina McMillan-Major, are co-authors on the paper On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜. As most of you undoubtedly know by now, there has been much controversy surrounding, and fallout from, this paper. In this conversation, our main priority was to focus on the message of the paper itself. We spend some time discussing the historical context for the paper, then turn to the goals of the paper, discussing the many reasons why the ever-growing datasets and models are not necessarily the direction we should be going. 

We explore the cost of these training datasets, both literal and environmental, as well as the bias implications of these models, and of course the perpetual debate about responsibility when building and deploying ML systems. Finally, we discuss the thin line between AI hype and useful AI systems, and the importance of doing pre-mortems to truly flesh out any issues you could potentially come across prior to building models, and much much more. 

The complete show notes for this episode can be found at twimlai.com/go/467.

Today we’re joined by Abhishek Gupta, a PhD Student at UC Berkeley. 

Abhishek, a member of the BAIR Lab, joined us to talk about his recent robotics and reinforcement learning research and interests, which focus on applying RL to real-world robotics applications. We explore the concept of reward supervision, and how to get robots to learn these reward functions from videos, and the rationale behind supervised experts in these experiments. 

We also discuss the use of simulation for experiments, data collection, and the path to scalable robotic learning. Finally, we discuss gradient surgery vs gradient sledgehammering, and his ecological RL paper, which focuses on the “phenomena that exist in the real world” and how humans and robotics systems interface in those situations. 

The complete show notes for this episode can be found at https://twimlai.com/go/466.

Today we’re joined by David Carmona, General Manager of Artificial Intelligence & Innovation at Microsoft. 

In our conversation with David, we focus on his work on AI at Scale, an initiative focused on the change in the ways people are developing AI, driven in large part by the emergence of massive models. We explore David’s thoughts about the progression towards larger models, the focus on parameters and how it ties to the architecture of these models, and how we should assess how attention works in these models.

We also discuss the different families of models (generation & representation), the transition from CV to NLP tasks, and an interesting point of models “becoming a platform” via transfer learning.

The complete show notes for this episode can be found at twimlai.com/go/465.

Today we’re joined by Melanie Mitchell, Davis Professor at the Santa Fe Institute and author of Artificial Intelligence: A Guide for Thinking Humans. 

While Melanie has had a long career with a myriad of research interests, we focus on a few, complex systems and the understanding of intelligence, complexity, and her recent work on getting AI systems to make analogies. We explore examples of social learning, and how it applies to AI contextually, and defining intelligence. 

We discuss potential frameworks that would help machines understand analogies, established benchmarks for analogy, and if there is a social learning solution to help machines figure out analogy. Finally we talk through the overall state of AI systems, the progress we’ve made amid the limited concept of social learning, if we’re able to achieve intelligence with current approaches to AI, and much more!

The complete show notes for this episode can be found at twimlai.com/go/464.

Today we’re joined by Adriana Kovashka, an Assistant Professor at the University of Pittsburgh.

In our conversation with Adriana, we explore her visual commonsense research, and how it intersects with her background in media studies. We discuss the idea of shortcuts, or faults in visual question answering data sets that appear in many SOTA results, as well as the concept of masking, a technique developed to assist in context prediction. Adriana then describes how these techniques fit into her broader goal of trying to understand the rhetoric of visual advertisements. 

Finally, Adriana shares a bit about her work on robust visual reasoning, the parallels between this research and other work happening around explainability, and the vision for her work going forward. 

The complete show notes for this episode can be found at twimlai.com/go/463.

Today we’re joined by Nishan Subedi, VP of Algorithms at Overstock.com.

In our conversation with Nishan, we discuss his interesting path to MLOps and how ML/AI is used at Overstock, primarily for search/recommendations and marketing/advertisement use cases. We spend a great deal of time exploring machine learning architecture and architectural patterns, how he perceives the differences between architectural patterns and algorithms, and emergent architectural patterns that standards have not yet been set for.

Finally, we discuss how the idea of anti-patterns was innovative in early design pattern thinking and if those concepts are transferable to ML, if architectural patterns will bleed over into organizational patterns and culture, and Nishan introduces us to the concept of Squads within an organizational structure.

The complete show notes for this episode can be found at https://twimlai.com/go/462.

Today we’re joined by Vered Shwartz, a Postdoctoral Researcher at both the Allen Institute for AI and the Paul G. Allen School of Computer Science & Engineering at the University of Washington.

In our conversation with Vered, we explore her NLP research, where she focuses on teaching machines common sense reasoning in natural language. We discuss training using GPT models and the potential use of multimodal reasoning and incorporating images to augment the reasoning capabilities.

Finally, we talk through some other noteworthy research in this field, how she deals with biases in the models, and Vered's future plans for incorporating some of the newer techniques into her future research.

The complete show notes for this episode can be found at https://twimlai.com/go/461. 

Today we’re joined by returning guest and newly appointed Dean of the College of Engineering at The Ohio State University, Ayanna Howard. 

Our conversation with Dr. Howard focuses on her recently released book, Sex, Race, and Robots: How to Be Human in the Age of AI, which is an extension of her research on the relationships between humans and robots. We continue to explore this relationship through the themes of socialization introduced in the book, like associating genders to AI and robotic systems and the “self-fulfilling prophecy” that has become search engines. 

We also discuss a recurring conversation in the community around AI  being biased because of data versus models and data, and the choices and responsibilities that come with the ethical aspects of building AI systems. Finally, we discuss Dr. Howard’s new role at OSU, how it will affect her research, and what the future holds for the applied AI field. 

The complete show notes for this episode can be found at https://twimlai.com/go/460.

Today we’re joined by returning guest and newly appointed Dean of the College of Engineering at The Ohio State University, Ayanna Howard. 

Our conversation with Dr. Howard focuses on her recently released book, Sex, Race, and Robots: How to Be Human in the Age of AI, which is an extension of her research on the relationships between humans and robots. We continue to explore this relationship through the themes of socialization introduced in the book, like associating genders to AI and robotic systems and the “self-fulfilling prophecy” that has become search engines. 

We also discuss a recurring conversation in the community around AI  being biased because of data versus models and data, and the choices and responsibilities that come with the ethical aspects of building AI systems. Finally, we discuss Dr. Howard’s new role at OSU, how it will affect her research, and what the future holds for the applied AI field. 

The complete show notes for this episode can be found at https://twimlai.com/go/460.

Today we’re joined by Penousal Machado, Associate Professor and Head of the Computational Design and Visualization Lab in the Center for Informatics at the University of Coimbra. 

In our conversation with Penousal, we explore his research in Evolutionary Computation, and how that work coincides with his passion for images and graphics. We also discuss the link between creativity and humanity, and have an interesting sidebar about the philosophy of Sci-Fi in popular culture. 

Finally, we dig into Penousals evolutionary machine learning research, primarily in the context of the evolution of various animal species mating habits and practices.

The complete show notes for this episode can be found at twimlai.com/go/459.  

Today we’re joined by Arul Menezes, a Distinguished Engineer at Microsoft. 

Arul, a 30 year veteran of Microsoft, manages the machine translation research and products in the Azure Cognitive Services group. In our conversation, we explore the historical evolution of machine translation like breakthroughs in seq2seq and the emergence of transformer models. 

We also discuss how they’re using multilingual transfer learning and combining what they’ve learned in translation with pre-trained language models like BERT. Finally, we explore what they’re doing to experience domain-specific improvements in their models, and what excites Arul about the translation architecture going forward. 

The complete show notes for this series can be found at twimlai.com/go/458.

Today we’re joined by Luna Dong, Sr. Principal Scientist at Amazon.

In our conversation with Luna, we explore Amazon’s expansive product knowledge graph, and the various roles that machine learning plays throughout it. We also talk through the differences and synergies between the media and retail product knowledge graph use cases and how ML comes into play in search and recommendation use cases. Finally, we explore the similarities to relational databases and efforts to standardize the product knowledge graphs across the company and broadly in the research community.

The complete show notes for this episode can be found at https://twimlai.com/go/457.

Today we’re joined by Sarah Brown, an Assistant Professor of Computer Science at the University of Rhode Island.

In our conversation with Sarah, whose research focuses on Fairness in AI, we discuss why a “systems-level” approach is necessary when thinking about ethical and fairness issues in models and algorithms. We also explore Wiggum: a fairness forensics tool, which explores bias and allows for regular auditing of data, as well as her ongoing collaboration with a social psychologist to explore how people perceive ethics and fairness.

Finally, we talk through the role of tools in assessing fairness and bias, and the importance of understanding the decisions the tools are making.

The complete show notes can be found at twimlai.com/go/456.

Today we’re joined by Andrew Trister, Deputy Director for Digital Health Innovation at the Bill & Melinda Gates Foundation. 

In our conversation with Andrew, we explore some of the AI use cases at the foundation, with the goal of bringing “community-based” healthcare to underserved populations in the global south. We focus on COVID-19 response and improving the accuracy of malaria testing with a bayesian framework and a few others, and the challenges like scaling these systems and building out infrastructure so that communities can begin to support themselves. 

We also touch on Andrew's previous work at Apple, where he helped develop what is now known as Research Kit, their ML for health tools that are now seen in apple devices like phones and watches.

The complete show notes for this episode can be found at https://twimlai.com/go/455

Today we’re joined by Drago Anguelov, Distinguished Scientist and Head of Research at Waymo. 

In our conversation, we explore the state of the autonomous vehicles space broadly and at Waymo, including how AV has improved in the last few years, their focus on level 4 driving, and Drago’s thoughts on the direction of the industry going forward. Drago breaks down their core ML use cases, Perception, Prediction, Planning, and Simulation, and how their work has lead to a fully autonomous vehicle being deployed in Phoenix. 

We also discuss the socioeconomic and environmental impact of self-driving cars, a few research papers submitted to NeurIPS 2020, and if the sophistication of AV systems will lend themselves to the development of tomorrow’s enterprise machine learning systems.

The complete show notes for this episode can be found at twimlai.com/go/454. 

Today we’re joined by Ya Xu, head of Data Science at LinkedIn, and TWIMLcon: AI Platforms 2021 Keynote Speaker.

We cover a ton of ground with Ya, starting with her experiences prior to becoming Head of DS, as one of the architects of the LinkedIn Platform. We discuss her “three phases” (building, adoption, and maturation) to keep in mind when building out a platform, how to avoid “hero syndrome” early in the process.

Finally, we dig into the various tools and platforms that give LinkedIn teams leverage, their organizational structure, as well as the emergence of differential privacy for security use cases and if it's ready for prime time.

The complete show notes for this episode can be found at https://twimlai.com/go/453. 

Today we’re joined by Jesse Engel, Staff Research Scientist at Google, working on the Magenta Project. 

In our conversation with Jesse, we explore the current landscape of creativity AI, and the role Magenta plays in helping express creativity through ML and deep learning. We dig deep into their Differentiable Digital Signal Processing (DDSP) library, which “lets you combine the interpretable structure of classical DSP elements (such as filters, oscillators, reverberation, etc.) with the expressivity of deep learning.”

Finally, Jesse walks us through some of the other projects that the Magenta team undertakes, including NLP and language modeling, and what he wants to see come out of the work that he and others are doing in creative AI research.

The complete show notes for this episode can be found at twimlai.com/go/452. 

Today we’re joined by return guest Francisco Webber, CEO & Co-founder of Cortical.io.

Francisco was originally a guest over 4 years and 400 episodes ago, where we discussed his company Cortical.io, and their unique approach to natural language processing. In this conversation, Francisco gives us an update on Cortical, including their applications and toolkit, including semantic extraction, classifier, and search use cases. We also discuss GPT-3, and how it compares to semantic folding, the unreasonable amount of data needed to train these models, and the difference between the GPT approach and semantic modeling for language understanding.

The complete show notes for this episode can be found at twimlai.com/go/451.

Today we’re joined by Gurdeep Pall, Corporate Vice President at Microsoft.

Gurdeep, who we had the pleasure of speaking with on his 31st anniversary at the company, has had a hand in creating quite a few influential projects, including Skype for business (and Teams) and being apart of the first team that shipped wifi as a part of a general-purpose operating system.

In our conversation with Gurdeep, we discuss Microsoft’s acquisition of Bonsai and how they fit in the toolchain for creating brains for autonomous systems with “machine teaching,” and other practical applications of machine teaching in autonomous systems. We also explore the challenges of simulation, and how they’ve evolved to make the problems that the physical world brings more tenable. Finally, Gurdeep shares concrete use cases for autonomous systems, and how to get the best ROI on those investments, and of course, what’s next in the very broad space of autonomous systems.

The complete show notes for this episode can be found at twimlai.com/go/450.

Today we’re joined by Bryan Carstens, a professor in the Department of Evolution, Ecology, and Organismal Biology & Head of the Tetrapod Division in the Museum of Biological Diversity at The Ohio State University.

In our conversation with Bryan, who comes from a traditional biology background, we cover a ton of ground, including a foundational layer of understanding for the vast known unknowns in species and biodiversity, and how he came to apply machine learning to his lab’s research.

We explore a few of his lab’s projects, including applying ML to genetic data to understand the geographic and environmental structure of DNA, what factors keep machine learning from being used more frequently used in biology, and what’s next for his group.

The complete show notes for this episode can be found at twimlai.com/go/449.

Today we’re joined by Jason Gauci, a Software Engineering Manager at Facebook AI.

In our conversation with Jason, we explore their Reinforcement Learning platform, Re-Agent (Horizon). We discuss the role of decision making and game theory in the platform and the types of decisions they’re using Re-Agent to make, from ranking and recommendations to their eCommerce marketplace.

Jason also walks us through the differences between online/offline and on/off policy model training, and where Re-Agent sits in this spectrum. Finally, we discuss the concept of counterfactual causality, and how they ensure safety in the results of their models.

The complete show notes for this episode can be found at twimlai.com/go/448.

Today we’re joined by Saiph Savage, a Visiting professor at the Human-Computer Interaction Institute at CMU, director of the HCI Lab at WVU, and co-director of the Civic Innovation Lab at UNAM.

We caught up with Saiph during NeurIPS where she delivered an insightful invited talk “A Future of Work for the Invisible Workers in A.I.”. In our conversation with Saiph, we gain a better understanding of the “Invisible workers,” or the people doing the work of labeling for machine learning and AI systems, and some of the issues around lack of economic empowerment, emotional trauma, and other issues that arise with these jobs.

We discuss ways that we can empower these workers, and push the companies that are employing these workers to do the same. Finally, we discuss Saiph’s participatory design work with rural workers in the global south.

The complete show notes for this episode can be found at twimlai.com/go/447.

Today we’re back with the final episode of AI Rewind joined by Michael Bronstein, a professor at Imperial College London and the Head of Graph Machine Learning at Twitter.

In our conversation with Michael, we touch on his thoughts about the year in Machine Learning overall, including GPT-3 and Implicit Neural Representations, but spend a major chunk of time on the sub-field of Graph Machine Learning. 

We talk through the application of Graph ML across domains like physics and bioinformatics, and the tools to look out for. Finally, we discuss what Michael thinks is in store for 2021, including graph ml applied to molecule discovery and non-human communication translation.

Today we continue the 2020 AI Rewind series, joined by friend of the show Sameer Singh, an Assistant Professor in the Department of Computer Science at UC Irvine. 

We last spoke with Sameer at our Natural Language Processing office hours back at TWIMLfest, and was the perfect person to help us break down 2020 in NLP. Sameer tackles the review in 4 main categories, Massive Language Modeling, Fundamental Problems with Language Models, Practical Vulnerabilities with Language Models, and Evaluation. 

We also explore the impact of GPT-3 and Transformer models, the intersection of vision and language models, and the injection of causal thinking and modeling into language models, and much more.

The complete show notes for this episode can be found at twimlai.com/go/445.

AI Rewind continues today as we’re joined by Pavan Turaga, Associate Professor in both the Departments of Arts, Media, and Engineering & Electrical Engineering, and the Interim Director of the School of Arts, Media, and Engineering at Arizona State University.

Pavan, who joined us back in June to talk through his work from CVPR ‘20, Invariance, Geometry and Deep Neural Networks, is back to walk us through the trends he’s seen in Computer Vision last year. We explore the revival of physics-based thinking about scenes, differential rendering, the best papers, and where the field is going in the near future.

We want to hear from you! Send your thoughts on the year that was 2020 below in the comments, or via Twitter at @samcharrington or @twimlai.

The complete show notes for this episode can be found at twimlai.com/go/444

Today we kick off our annual AI Rewind series joined by friend of the show Pablo Samuel Castro, a Staff Research Software Developer at Google Brain.

Pablo joined us earlier this year for a discussion about Music & AI, and his Geometric Perspective on Reinforcement Learning, as well our RL office hours during the inaugural TWIMLfest. In today’s conversation, we explore some of the latest and greatest RL advancements coming out of the major conferences this year, broken down into a few major themes, Metrics/Representations, Understanding and Evaluating Deep Reinforcement Learning, and RL in the Real World.

This was a very fun conversation, and we encourage you to check out all the great papers and other resources available on the show notes page.

Today we close out our NeurIPS series joined by Aravind Rajeswaran, a PhD Student in machine learning and robotics at the University of Washington.

At NeurIPS, Aravind presented his paper MOReL: Model-Based Offline Reinforcement Learning. In our conversation, we explore model-based reinforcement learning, and if models are a “prerequisite” to achieve something analogous to transfer learning. We also dig into MOReL and the recent progress in offline reinforcement learning, the differences in developing MOReL models and traditional RL models, and the theoretical results they’re seeing from this research.

The complete show notes for this episode can be found at twimlai.com/go/442

As we continue our NeurIPS 2020 series, we’re joined by friend-of-the-show Charles Isbell, Dean, John P. Imlay, Jr. Chair, and professor at the Georgia Tech College of Computing.

This year Charles gave an Invited Talk at this year’s conference, You Can’t Escape Hyperparameters and Latent Variables: Machine Learning as a Software Engineering Enterprise. In our conversation, we explore the success of the Georgia Tech Online Masters program in CS, which now has over 11k students enrolled, and the importance of making the education accessible to as many people as possible. We spend quite a bit speaking about the impact machine learning is beginning to have on the world, and how we should move from thinking of ourselves as compiler hackers, and begin to see the possibilities and opportunities that have been ignored.

We also touch on the fallout from Timnit Gebru being “resignated” and the importance of having diverse voices and different perspectives “in the room,” and what the future holds for machine learning as a discipline.

The complete show notes for this episode can be found at twimlai.com/go/441. 

Today we kick off our NeurIPS 2020 series joined by Taco Cohen, a Machine Learning Researcher at Qualcomm Technologies.

In our conversation with Taco, we discuss his current research in equivariant networks and video compression using generative models, as well as his paper “Natural Graph Networks,” which explores the concept of “naturality, a generalization of equivariance” which suggests that weaker constraints will allow for a “wider class of architectures.”

We also discuss some of Taco’s recent research on neural compression and a very interesting visual demo for equivariance CNNs that Taco and the Qualcomm team released during the conference.

The complete show notes for this episode can be found at twimlai.com/go/440.

Today we close out our re:Invent series joined by Edgar Bahilo Rodriguez, Lead Data Scientist in the industrial applications division of Siemens Energy.

Edgar spoke at this year's re:Invent conference about Productionizing R Workloads, and the resurrection of R for machine learning and productionalization. In our conversation with Edgar, we explore the fundamentals of building a strong machine learning infrastructure, and how they’re breaking down applications and using mixed technologies to build models.

We also discuss their industrial applications, including wind, power production management, managing systems intent on decreasing the environmental impact of pre-existing installations, and their extensive use of time-series forecasting across these use cases.

The complete show notes can be found at twimlai.com/go/439.