AI’s Progression ♟️ 1997 Deep Blue: used brute‑force search for best probabilistic moves. No learning. beat Kasparov. ⚪️ 2016 AlphaGo: neural nets learned via self‑play and beat Lee Sedol. Move 37 was creative by AI. Lee’s move 78 was 1 in 10,000 and broke AlphaGo and he won one game. Lost tournament. 🧮 2025 LLMs: read math texts and verify their own proofs and earn gold on IMO problems.