Five-Year Retrospective Argues ML-Assisted Directed Evolution Ignores Real-World Costs
anshulkundaje · x · 2026-09-06
Bruce Wittmann published a five-year retrospective on arXiv examining why machine learning has transformed much of protein engineering but not directed evolution. He argues the core problem is misaligned objectives: MLDE researchers optimize for 'identify an optimal protein' while practitioners need 'a sufficient protein under time and resource constraints.' He highlights that nearly all current MLDE methods ignore DNA synthesis costs, limiting practical applicability regardless of model quality, and points to recent exceptions while arguing the proposed reframing is compatible with existing ML progress.
More from Research
- OpenAI releases data on models accelerating research, urging industry transparency on RSI — kliu128 · 2026-09-06
- Researcher calls on RL teams to add refactoring and deletion tasks to agent training — kuza55 · 2026-09-06
- The 1958 perceptron was misunderstood both ways: NYT hype, then 11 years as a dead end — techNmak · 2026-09-06
- dair-ai weekly picks: Declarative Attention tops the week's AI papers — dair_ai · 2026-09-06
- Why Preserving a Pattern Likely Isn't Enough to Produce Consciousness — pwlot · 2026-09-06
- Consciousness debate: digital simulation can't instantiate causal physics, Church-Turing won't save you — pwlot · 2026-09-06