Perpetual Chess Podcast

EP 474- NM Han Schut on GM Jan Timman’s Legacy, Chess Improvement, & Using LLM’s to Create Helpful Chess Tools

Mar 3, 2026
Han Schut, National Master and Chessable author known for the Steps Method and chess tech, reflects on GM Jan Timman’s legacy. He recounts using Claude to rebuild searchable puzzle and opening tools, discusses creating AI-assisted chess study aids, and shares practical tips on managing fatigue, mindset, and effective puzzle practice.
Ask episode
AI Snips
Chapters
Books
Transcript
Episode notes
INSIGHT

Conversational Coding Unlocks Chess Tool Creation

  • Claude-style LLMs let non-programmers specify tools conversationally and produce working code for complex tasks.
  • Han used Claude to convert the Lichess puzzle dump into a searchable PGN of 1.2M opening puzzles within hours of iterative prompts.
ADVICE

Search Puzzles From Any Position

  • Query the opening-puzzle PGN from any exact board position to discover themes, frequencies, and vulnerabilities.
  • Han truncates games at the puzzle moment and stores the solution in the header for fast reference in ChessBase.
ADVICE

Spend $20 To Get A Usable LLM For Coding

  • Pay for the $20 Claude tier to get coding-capable models before experimenting.
  • Han recommends that modest subscription as the minimal investment to access reliably useful conversational coding power.
Get the Snipd Podcast app to discover more snips from this episode
Get the app