Andriy Burkov ships Hundred-Page Deep RL book, built ChapterPal with zero code via vibe coding
burkov · x · 2026-11-10
- Andriy Burkov (The Hundred-Page Machine Learning Book) announces his new Hundred-Page Deep Reinforcement Learning Book, covering policy gradients through PPO/DQN with complete PyTorch implementations and Jupyter notebooks.
- After resigning in Feb 2025 to write full-time, he vibe-coded ChapterPal, an AI-assisted reading platform, without writing a single line of code — fighting hallucinated and overwritten code along the way.
- Once ChapterPal's maintenance was automated enough last spring, he returned to writing. The book is endorsed by Lambda CTO Stephen Balaban and OpenAI's Long Ouyang.
More from coding & agent
- Asking Fable 5.2 to upload one static HTML page took a 3m52s headless-Chromium detour — victormustar · 2026-11-11
- Google cut type-checking peak compute by 80% after switching its Python monorepo from Pytype to Pyrefly — tdhopper · 2026-11-11
- TCS built a BFSI agentic test orchestrator on Bedrock AgentCore with per-tenant isolation — AWS ML Blog · 2026-11-11
- Fable 5.2 real-work review: success rate fell from 73% to 45% in three weeks, now supervising its own fix — leebase65 · 2026-11-11
- AWS official guide: build a Bedrock invocation observability data lake in five stages — AWS ML Blog · 2026-11-11
- Pichai demos Antigravity /swarm: subagents iteratively build a 3D pelican biking game — sundarpichai · 2026-11-11