I know Kung Fu! RubyGems for AI
summary
Claim: an LLM does not beat a human getting an app from idea to production; experiments target the gaps by wisdom gem that captures chat learnings into git. Tools: a Rails web API, a CLI gem that talks to OpenAI and Claude, Postgres with PGVector + git RAG for retrieval, plus MCP ideas for live repo inspection. Built outcome: demo stores preferences, indexes commits, and then retrieves context to fix a repeated nano annoyance — proof of concept works though full sharing hub and docs remain unfinished.
AI-written summary · about
chapters (8)
- 0:00AI and RubyGems intro
- 2:27Capturing AI learning
- 6:48Automating chat storage
- 9:38Git for AI context
- 12:11Demo with Wisdom gem
- 14:20Open source future
- 20:46Testing AI prompts
- 24:44Ruby/Rails for AI
related talks
35mThe State of the Ruby AI Toolbox
28mKeynote: Ruby Is the Best Language for Building AI Web Apps
01h 04mBuilding LLM powered applications in Ruby
21mRubyLLM
28mAPIs and GenAI : Your Perfect Combo with LLM-Powered Frameworks.
31mDefying Gravity: Teaching AI to Write Better Ruby
47mGlimpses of Humanity: My Game-Building AI Pair
30mLegacy Rails and the AI That Couldn't