www.daniellitt.com/blog/2026/9/13/a-beg…
"I think so. This machine might produce answers we value, but it would not, in itself, produce human understanding of those answers. In fact I think we are at the beginning of an incredible, wonderful explosion of mathematics, and if we value human understanding, there will be more need for human mathematicians than ever before. But the profession will have to change."
"I propose the following reconceptualization of the goal of a mathematics PhD: to become a world expert on some interesting, deep topic, and to be able to convey that interest and understanding to others. Part of operationalizing this might be a thesis, but the degree would be awarded primarily on the basis of a rigorous defense, in which the student explains the topic to their examiners until they are satisfied. While we might require the topic to be original, its provenance—AI or not—is irrelevant."
vidya 
naresh 
anand 
bhavesh Quite a nice read on how the future looks.