An OpenAI model has disproved a central conjecture in discrete geometry

To me, this result and other recent developments leave no doubt that state of the art AI is capable of highly innovative, original work, a position which I would have opposed a year ago. If you are still a skeptic, I recommend reading the commentary in the post from the multiple leading mathematicians in the fields that the proof touches on.

In spite of the many (valid) concerns surrounding AI, I believe that one must admit we are in a strange new world.