Jeff Dean just left Google after 27 years to found Discovery Loop to use AI to accelerate science and engineering research. AI will speed up drug discovery. It does not (yet) speed up making 400,000 doses and trucking them across a complex supply chain. As I think about the next bottlenecks in AI, I think about the real world. I'm thinking about atoms, not software. Physical, tangible, gets-stuck-in-a-forklift stuff. In addition to cleaner and more effective energy sources, we should be looking at companies that have a headstart in manufacturing and making factories faster. Looking at Amazon and large-scale drug makers (Novartis, Pfizer, AZ, GSK, Novo Nordisk, J&J, Merck, Eli Lilly, Roche) obviously. Also at places like Flexe for warehousing, Mytra for simplification of storage, Autostore for warehouse optimization, and Berkshire Grey. This is not investment advice. Next decade goes to whoever can produce more (or more effectively) in the physical world.