@AndrewYNg
I'd love to see the "AI for Science" community generate dozens of ideas that could have the impact of AlphaFold2, and identify enablers for progress, e.g. better defined problems, open datasets, benchmarks, high-throughput experimentation, automation,