HackerNews AIOctober 1, 2026Clef: Open-weight decision models, and new RL fine-tuning platformRead original article →