Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
Autolearn
Learn how Autolearn, a coding agent on an MCP server, captures reasoning steps and converts them into functional code, enabling agents to self‑construct.
A coding agent behind an MCP server that can “crystallize” reasoning steps into code
AutoLearn crystallizes AI reasoning into deterministic code skills, reducing inference costs.
FastAPI MCP server crystalizes AI reasoning into persistent, executable Python skills.
Related projects
Who Watches The Watchbots?
Seattle
We'll demonstrate a local‑first open‑source system that enforces verified AI control strategies, preventing assistants from bypassing tests or…
AI Agents playing Texas Hold'em
Seattle
A prototype demo of multiple AI agents playing Texas Hold'em, covering design, implementation hurdles, and recommended frameworks and…
Dev Rel automation
Seattle
Learn how we built an automated Reddit response system that generates accurate tone and syntactically correct code for…
AI Startup Scout
Seattle
A sub‑agent scans Hacker News, flags potential startup stories, adds founder data from external sources, and generates concise…
Agentic Control of Research Labratories (AI Scientists!)
San Francisco
Demonstration of multi-agent systems controlling lab automation to design, execute, and review wet‑lab experiments, including deployment in a…
Recursive Agentic builder - Build an agent with a prompt.
Seattle
Learn how a CLI uses recursive reinforcement learning to generate, test, and refine functional AI agents from a…