There's something deeply philosophical about the importance of human understanding in a world where most work is done by AI.

Terence Tao's whole point in the letter isn't that AI is universally bad, but that it should be used in a way that complements the advancement of human understanding rather than doing work (like solving math proofs) for its own sake.

You can obviously use AI to help accelerate research and understanding, but the point here is that you also don't have to in order to get things done.

As someone who's largely optimistic about AI, I don't have concrete opinions on this matter yet, but I empathize with the broader question of how deeply humans should strive to understand the world vs. fully offloading it to others.

Andrew Curran (@AndrewCurran_) Twenty-five Fields Medal winners have published a joint declaration warning about what they see as a severe misalignment between AI companies and the mathematics community.