Episode 24: AI + Hacking with Daniel Miessler and Rez0

Episode 24: In this episode of Critical Thinking - Bug Bounty Podcast, we chat with Daniel Miessler and Rez0 about the emergence and potential of AI in hacking. We cover AI shortcuts and command line tools, AI in code analysis and the use of AI agents, and even brainstorm about the possible opportunities that integrating AI into hacking tools like Caido and Burp might present. Don't miss this episode packed with valuable insights and cutting-edge strategies for both beginners and seasoned bug bounty hunters alike.

Follow us on twitter at: @ctbbpodcast

We're new to this podcasting thing, so feel free to send us any feedback here: [email protected]

Shoutout to YTCracker for the awesome intro music!

------ Links ------

Follow your hosts Rhynorater & Teknogeek on twitter:

https://twitter.com/0xteknogeek

https://twitter.com/rhynorater

Today’s Guests:

https://twitter.com/rez0__

https://twitter.com/DanielMiessler

Daniel Miessler’s Unsupervised Learning

https://danielmiessler.com/

Simon Willison's Python Function Search Tool

https://simonwillison.net/2023/Jun/18/symbex/

oobabooga - web interface for models

https://github.com/oobabooga/text-generation-webui

State of GPT

https://karpathy.ai/stateofgpt.pdf

AI Canaries

https://danielmiessler.com/p/ai-agents-canaries

GPT3.5

https://community.openai.com/t/gpt-3-5-turbo-0613-function-calling-16k-context-window-and-lower-prices/263263

GPT Engineer

https://github.com/AntonOsika/gpt-engineer

Timestamps:

(00:00:00) Introduction

(00:05:40) Using AI for hacking: Developing hacking tools and workflow shortcuts

(00:11:40) GPT Engineer and Small Developer for Security Vulnerability Mapping

(00:22:40) The potential dangers of centralized vs. decentralized finance

(00:24:10) Ethical hacking and circumventing ChatGPT restrictions

(00:26:09) AI Agents, Reverse API, and Encoding/Decoding Tools

(00:31:45) Limitations of AI in context window and processing large JavaScript files

(00:36:50) Meta-prompter: Enhancing prompts for accurate responses from GPT

(00:41:00) GPT-35 and the new 616K context model

(45:08) Creating a loader for Burp Suite files or Caido instances

(00:54:02) Hacking AI Features: Best Practices

(01:00:00) AI plugin takeover and the need for verification of third-party plugins and tools