For years, the idea sounded improbable: write a molecular shape into existence, then persuade nature to build it. Baker kept at it, and built a community that could try.

He started with a stubborn software problem: brilliant protein-design tools were too hard to use. A decade later, Lucas Nivon is still widening the doorway - now between algorithms, experiments, and open scientific infrastructure.
The Seattle spinout made Rosetta easier for pharma scientists, learned from more than 30 partnered discovery programs, and then made a riskier bet: building its own autoimmune drugs. Its story is a useful lesson in when software should become a laboratory - and when that strategy can still fail.