Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
Frontier Evals and System Hacking
This talk covers evaluating frontier AI models and demonstrates techniques for identifying and exploiting vulnerabilities in a simulated system environment.
Frontier model evals & hacking a vulnerable system
Related projects
Coding agents and data science: building hill-climbing environments for LLMs
London
Dozens of coding agents attempt to reverse engineer a Spotify color assignment algorithm within a custom environment, showcasing…
Building Conversational AI Agents
London
Learn practical steps to design, develop, and deploy conversational AI agents, covering architecture, language models, training data, evaluation,…
Unified AI Rules Management: How to Prevent Vendor Lock-In Across AI Coding Tools
Denver
See a live demo of rulesync, a tool unifying AI coding assistant rules across Claude Code, Copilot, Cursor,…
Financial Crime Detection at Monzo
London
This talk explores methods and challenges in detecting financial crime at Monzo, focusing on practical approaches and real-world…
How I hacked the hottest SF startup
Poland
Revealing how Poke's system prompt was leaked and its architecture reverse-engineered, this talk demonstrates near-future AI interaction via…
A Fireside chat with Guy Podjarny, founder of Tessl, on AI Native Development
London
Explore how AI shifts software development from code‑centric to spec‑centric, letting users define desired outcomes while AI handles…