@fogti I guess I'm not really making a very strong (in the sense of big claims) argument, but I do think large models will have both a direct impact on medicine/molecular biology and an indirect one through the innovation they're spurring in both techniques and hardware.

There have also been some pretty significant advances in molecular biology over the past few years that aren't directly AI related but act as "multipliers", in particular gene sequencing and synthetic biology.

Synthetic biology also has uses far beyond medicine. For example, it seems likely that it will allow us to produce "biofuels" that are chemically identical to fuels that we already use, thus rendering the notion of a "hydrogen economy" (which is fossil fuel industry propaganda IMO) irrelevant.