This is a valid take but at this moment, AI is not trying to solve this fundamental dynamic. But it can still accelerate the process by aggressive exploration o

This is a valid take but at this moment, AI is not trying to solve this fundamental dynamic. But it can still accelerate the process by aggressive exploration of the solution space which cannot be done even with an army of human researchers. Many ideas can be relatively quickly verified (and discarded if needed) by proper simulation even before real world experimentation, but we don't have enough capacity to process all potential ideas. If you can build a good model for simulation and establish a robust methodologies, we can use some ideas which never had a chance before.