I’ve talked to a few people about creating a compute fund where the idea is we would sponsor the ai credits or compute resources for anyone who has a new promising approach to important problems in math, physics, & engineering. It’s very clear that right now this is the highest roi application of AI. Some of you might be skeptical since you can only get so far with theory but I’m reminded of how much impact Terrance Tao had with just his work on compressed sensing which leads me to believe we probably have $1 trillion dollars of low hanging fruit left before we hit the experimental bottlenecks. Dm if interested. Also would love to give the caltech kids a head start !
