What's this about?
Sequoia's Training Data hosts Chai Discovery, who argue that designing molecules and drugs is fundamentally a scaling problem — throw enough data and compute at general methods and they beat hand-crafted approaches.
What exactly is "the Bitter Lesson"?
A famous essay by Rich Sutton: in AI, general methods that leverage lots of computation tend to win over clever human-designed shortcuts. Applying it to drug design means betting on scalable learning over decades of specialized biochemistry heuristics.
Why it matters
If the Bitter Lesson holds in biology, it reshapes how (and how fast) new drugs get discovered.