Mixture of Experts

What is all the hype around Deep Research? In episode 43 of Mixture of Experts, host Tim Hwang is joined by Kate Soule, Volkmar Uhlig and Shobhit Varshney. This week, we discuss reasoning model features coming out of companies like OpenAI’s Deep Research, Google Gemini, Perplexity, xAI’s Grok-3 and more! Next, OpenAI is rumored to release an inference chip, but how likely is this to be a success in the AI chip game? Then, we analyze the capabilities of small vision-language models (VLMs). Finally, a startup, Firecrawl, released a job posting in search of an AI agent. Is this the future for AI tools in the workforce? Tune-in to today’s Mixture of Experts to find out.

00:01 – Intro

00:35 – Deep Research

11:58 – OpenAI inference chip

22:17 – Small VLMs

32:31 – AI agent job posting

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Live from Paris, Tim Hwang is at the AI Action Summit 2025. In episode 42 of Mixture of Experts, we welcome Anastasia Stasenko, CEO and Co-Founder of pleias along with our veteran experts Marina Danilevsky and Chris Hay. Last week, we touched on some potential conversations at the Paris AI Summit, this week we recap what actually happened. Is AI safety improving Globally? Next, for our paper of the week, we breakdown s1: Simple test-time scaling. Then, Sam Altman is back with another blog, “Three Observations,” what do our experts have to say? Finally, what can we learn from Anthropic’s Economic Index? All that and more on today’s Mixture of Experts. 

00:01 – Intro 

00:42 – Paris AI Summit 

11:10 – s1: Simple test-time scaling 

19:32 – Sam Altman’s “Three Observations” 

30:41 – Anthropic’s Economic Index 

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

Resources:

Read the paper about s1: Simple test-time scaling: https://arxiv.org/abs/2501.19393

Read Sam Altman's "Three Observations": https://blog.samaltman.com/three-observations

Read Anthropic's Economic Index: https://www.anthropic.com/economic-index

Read more about AGI: https://www.ibm.com/think/topics/artificial-general-intelligence

What does Sam Altman have up his sleeve? In episode 41 of Mixture of Experts, join host Tim Hwang along with experts Nathalie Baracaldo, Marina Danilevsky and Chris Hay. Last week, we covered all things DeepSeek, and this week OpenAI has some new releases to share. Today, the experts dissect deep research and o3-mini. Next, our host Tim Hwang is travelling to AI Action Summit, he asks our experts what we can expect coming out of the event. Then, we talk about Anthropic’s Constitutional Classifiers. Finally, Microsoft is creating a unit to study AI’s impact, what does this mean? Find out all this and more on Mixture of Experts.

00:01 – intro

00:41 – Open AI deep research and o3-mini

13:51 – AI Action Summit

20:17 – Anthropic’s Constitutional Classifiers

28:54 – Microsoft AI Impact team

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

• Subscribe for AI updates
• Learn more about artificial intelligence
• DeepSeek's reasoning AI shows power of small models, efficiently trained
• Visit Mixture of Experts podcast page to learn more AI content

Let’s bust some early myths about DeepSeek. In episode 40 of Mixture of Experts, join host Tim Hwang along with experts Aaron Baughman, Chris Hay and Kate Soule. Last week, we covered the release of DeepSeek-R1; now that the entire world is up to speed, let’s separate the facts from the hype. Next, what is model distillation and why does it matter for competition in AI? Finally, Sam Altman among other tech CEOs shared his response to DeepSeek. Will R1 radically change the open-source strategy of other tech giants? Find out all this and more on Mixture of Experts.

00:01 – Intro

00:41 – DeepSeek facts vs hype

21:00 – Model distillation

31:21 – Open source and OpenAI

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

What does the future hold for DeepSeek? In episode 39 of Mixture of Experts, join host Tim Hwang along with experts Abraham Daniels, Kaoutar El Maghraoui and Skyler Speakman to discuss the release of DeepSeek-R1. Next, Mistral indicates going IPO. Then, FrontierMath’s new benchmark is particularly difficult, the experts debrief. Finally, IDC released a report on code assistants, what do we need to know about generalist and specialized coding assistants? Tune-in to this week’s episode to find out. 

00:01 – Intro  

01:08 – DeepSeek-R1 

14:08 – Mistral indicates IPO 

20:54 – FrontierMath controversy 

30:04 -- IDC code assistants report 

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

What would you do with $2 billion? In episode 38 of Mixture of Experts, join host Tim Hwang along with experts Chris Hay, Kaoutar El Maghraoui and Vyoma Gajjar to discuss the Anthropic valuation rumors. Next, Microsoft CEO Nadella created a new CoreAI group to build and run apps for customers. Then, NotebookLM upgraded some of its features, including podcast intervention. Finally, AI agents are making their way into the financial services industry. Can an agent invest all of your money? Tune-in to this week’s episode to find out. 

00:01 -- What would you do with $2 billion? 

00:51 -- Anthropic valuation 

12:14 -- Microsoft CoreAI 

25:01 -- NotebookLM upgrades 

35:17 -- AI agents in finance 

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

What’s the most exciting CES AI announcement? In episode 37 of Mixture of Experts, host Tim Hwang is joined by Skyler Speakman, Volkmar Uhlig and Shobhit Varshney to debrief CES 2025. Specifically, the experts dive into NVIDIA’S Project DIGITS, among other announcements from the AI hardware giant. Next, a new enterprise AI development survey came out that detailing how developers really feel about AI implementation. Then, Apple Intelligence experienced some major hallucination fails, what does this tell us about Apple’s stake in the AI game? Finally, Sam Altman of OpenAI released a reflection blog, what does he say about the future of AI? All that and more on today’s Mixture of Experts.

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Is deep learning hitting a wall? It’s 2025 and Mixture of Experts is back and better than ever. In episode 36, host Tim Hwang is joined by Chris Hay, Kate Soule and Kush Varshney to debrief one of the biggest releases of 2024, OpenAI o3. Next, DeepSeek-V3 is here! Finally, will AI exist in 2027? The experts dissect the AI bet between Miles Brundage and Gary Marcus. All that and more on the first Mixture of Experts of 2025.

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

00:00 — Intro

00:49 — OpenAI o3

14:40 — DeepSeek-V3

28:00 — The Brundage/Marcus bet

Will 2025 be the year of AI agents? In Episode 35 of Mixture of Experts, host Tim Hwang is joined by some show veterans to debrief 2024 in AI. This week, we review AI models, agents, hardware and product releases with some of the top industry experts. What was the best model of 2024? Is NVIDIA king? What are some of the AI trends in 2025? All that and more on this special edition of Mixture of Experts.

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Is pre-training a thing of the past? In Episode 34 of Mixture of Experts, host Tim Hwang is joined by Abraham Daniels, Vagner Santana and Volkmar Uhlig to debrief this week in AI. First, OpenAI cofounder Ilya Sutskever said that “peak data” was achieved, does this mean there is no longer a need to model pre-training? Next, IBM released Granite 3.1 with a slew of features, we cover them all. Then, there is a new way to steal AI models, how do we protect against model exfiltration. Finally, can NVIDIA Jetson for AI developers really increase hardware accessibility? Tune-in for more!

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

00:01 — Intro

00:49— Is pre-training over?

10:25 — Granite 3.1

22:23 — AI model stealing

33:38—NVIDIA Jetson

Is o1 Pro worth the cost? In Episode 33 of Mixture of Experts, host Tim Hwang is joined by Marina Danilevsky, Kate Soule and Vyoma Gajjar. First, the experts debrief the 12 Days of OpenAI. Next, we review some of the top papers in NeurIPS, how are the experts keeping up with all these research papers? Then, we are back with another benchmark, can ARC Prize make AGI more tractable? Finally, Meta announced the launch of Llama 3.3 70B with the promise of 405B performance, can we have our cake and eat it too? Find out more on today’s Mixture of Experts!

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

What’s the mystery behind the name ChatGPT refuses to discuss? In Episode 32 of Mixture of Experts host Tim Hwang dives into the hottest topics shaping the AI landscape with an all-star panel: Aaron Baughman, Vagner Figueredo de Santana, and Shobhit Varshney. First, they disect the biggest announcements and takeaways from AWS re:Invent 2024, Amazon’s premier AI event. Next, they talk about overcoming architectural vulnerabilities in AI systems, and finally, they uncover the curious case of a name ChatGPT won’t discuss—and the questions this raises about privacy and transparency in AI. Get ready for an episode packed with insights, debates, and forward-thinking perspectives!

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

How much future learning will be done with an AI assistant? In Episode 31 of Mixture of Experts, host Tim Hwang is joined by Phaedra Boinodiris, Marina Danilevsky and Skyler Speakman for the AI in education special episode. First, the experts give an update on the state of AI within education. Next, we cover concerns around AI safety and literacy, what do students and teachers need to be aware of? Finally, the panel gives their predictions on what the future of education holds as it relates to AI. Tune-in to this special episode for an in-depth analysis!

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Should your AI assistant remember everything about you? In Episode 30 of Mixture of Experts, host Tim Hwang is joined by Vagner Santana, Vyoma Gajjar and Shobhit Varshney. First, the experts breakdown claims of “near-infinite memory” within AI models. Next, Shobhit is fresh off the plane from Microsoft Ignite, he shares some of the exciting new announcements following the event. Then, a new benchmark has entered the chat, what do we know about FrontierMath? Finally, AlphaFold3 is now more open, why does this matter? Find out more on today’s episode!

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Is 2024 the year scaling AI officially breaks? In Episode 29 of Mixture of Experts, host Tim Hwang is joined by Anthony Annunziata, Kate Soule and Naveen Rao. First, the experts discuss whether we are living in a post scale world. Next, we can’t have an episode without chatting AI agents, but what does the future hold for this technology? Finally, is AGI here to stay? Tune-in to this week’s Mixture of Experts to find out.

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Could AI wipe out software engineers? In Episode 28 of Mixture of Experts, host Tim Hwang is joined by Chris Hay, Kaoutar El Maghraoui, and Shobhit Varshney. First, the experts discuss GitHub reporting a rise of developers driven by AI code assistant tools. Next, Big Sleep finds a vulnerability in SQLite, what is the future for these kinds of AI agents? Finally, OpenAI released SearchGPT, what is the future of AI search? Tune-in today to find out!

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Agents, agents, and more agents! In Episode 27 of Mixture of Experts, host Tim Hwang is joined by Volkmar Uhlig and Vyoma Gajjar. First, the experts chat about Mark Benioff’s spicy tweet, and what this means for the future of AI agents. Next, how much energy is needed to power AI models, and should we be concerned? Then, the experts debrief Anthropic’s release of computer use. Finally, Google is integrating SynthID-Text into Gemini to help watermark AI-generated text, do we need this feature? Learn more on today’s Mixture of Experts.

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Can chat replace search? In Episode 26 of Mixture of Experts, host Tim Hwang is joined by Kate Soule, Kush Varshney and Petros Zerfos for IBM TechXchange week! First, the experts describe how the team created the Granite 3.0 models. Next, NVIDIA enters the open source model game, what does this mean for the competition? Finally, Perplexity AI is seeking over double their valuation in new funding rounds, what does this mean for start-ups? All that and more on today’s episode.

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Can AI solve infectious disease? In Episode 25 of Mixture of Experts, host Tim Hwang is joined by Kaoutar El Maghraoui, Maya Murad, and Ruben Boonen. Today we analyze some papers. First, the experts dissect Machines of Loving Grace, a 15,000 word essay written by Anthropic’s CEO making some major AI predictions. Then, Apple generated a new benchmark based of GSM8K in a recent paper, the findings were intriguing. Next, we talk Entropix, a sampler intending to replicate chain of thought features. Finally, OpenAI disclosed they are seeing an increase in AI models faking articles, what can we do to fix this? All this and more, on today’s Mixture of Experts.

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Could AI win a Nobel Prize in the future? In Episode 24 of Mixture of Experts, host Tim Hwang is joined by Chris Hay and Edward Calvesbert. First, the experts debrief the ‘Godfather of AI’ sharing a Nobel Prize. Next, we talk AI platforms and the hype around DGX B200. Finally, unstructured data is becoming usable for LLMs, why are companies like NVIDIA so interested in this data? Tune-in today to find out!

The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.