Google DeepMind releases AlphaGenome 2 for drug target discovery
Benchmark performance nearly doubled against its predecessor, which matters more for drug discovery teams than raw model size comparisons. Genomics researchers and biotech teams can use variant-to-cell-type predictions to prioritise which gene targets to validate experimentally. Compared with frontier language model releases dominating headlines, this shows AI value concentrating in scientific tooling, a direction enterprise R&D leaders evaluating AI investments should watch closely.
