Tom Gebhart

Machine learning models now outperform the best numerical weather prediction systems in both speed and accuracy. But the theory underlying their impressive performance is as old as numerical weather prediction itself.

Large language models are learning to code, but can they reason well enough to produce novel algorithms that can solve combinatorial optimization problems?

Given the rapid pace of progress within natural language processing, it seems like we should be able to automate the generation of new high-quality scientific hypotheses. But if this were possible, would we ever hear about it?