A year ago, I would not have expected the first academic field to seem to reach a consensus that AIs will accelerate research (which is not the same thing as autonomous research) would be math But that appears to be happening based on math professors in my feed and elsewhere.

Some examples