Predict hidden reaction rates, not yields: rethinking AI models for chemical discovery
bravo_abad · x · 2026-09-24
Chemistry researcher Jorge Bravo Abad argues on Substack that in nickel cross-coupling systems, aryl and alkyl partners activate through different pathways — reactions can fail even when every step is chemically plausible, because side reactions win in the wrong kinetic regime. His proposal: instead of training models directly on reaction yield, predict the hidden underlying rates, making kinetics the modeling target. A thoughtful take on choosing the right modeling object in AI-driven scientific discovery.
Related event: ML Maps Kinetics of Nickel-Catalyzed Coupling Instead of Yield(2 posts)→
More from Research
- Benchmark run finds "arjunomics in the weights", raising misalignment concerns — kenbwork · 2026-09-24
- Arena opens Fall 2026 academic program with $50K per AI evaluation project — arena · 2026-09-24
- RSA-896 Factored With Claude Orchestrating 2,048 GPUs Over 10 Days, 30 GPU-Years — matthew_d_green · 2026-09-24
- Dario touts Claude-led gene editing discovery, but biologists call it overhyped — ccerrato147 · 2026-09-24
- Jev Is an Architecture Bet, Not a Product: Could LLMs Specialize Beyond Generation? — tenkei_01 · 2026-09-24
- Mathematician proposes nonlinear variant of Komlós conjecture, inspired by AI proof — burny_tech · 2026-09-24