Granite Asia sees lab testing as bottleneck in AI-driven drug discovery
XLV•Granite Asia says laboratory testing is becoming a bottleneck in AI-driven drug discovery, shifting the focus toward experimental validation. Earendil Labs says it can prepare molecules for manufacturing and toxicology studies within one year, compared with about three, while its CEO estimates autonomous, self-improving life-sciences models remain three to five years away.
1. Testing and model development
Laboratory testing is becoming a bottleneck in AI-driven drug discovery, making reliable experimental validation more important alongside molecule design. Earendil Labs says AI design and high-throughput biology can prepare molecules for manufacturing and toxicology studies within one year, versus about three. BioMap uses customers’ historical experimental data to build task-specific models. Earendil CEO Peng Jian estimates autonomous self-improving models in life sciences remain three to five years away, with data accumulation still underway.
