Karthik C. S., Amit Sahai, and Jie Xue argue that AI-assisted mathematical discovery exposes a problem with measuring intellectual worth through individual achievement. Their guest essay on Terence Tao’s blog locates that worth in understanding a truth, whoever or whatever discovers it.
- Correctness and insight differ. Producing valid deductions leaves open which questions matter and why a result illuminates anything.
- Human superiority is a fragile foundation. Defending creativity or intuition mainly because humans still outperform machines preserves the same competitive measure of worth.
- Understanding can be shared. Someone else’s discovery need not diminish its beauty or the pleasure of learning it.
The authors offer an ethical orientation and explicitly leave policy unsettled. Both using AI and choosing to abstain can express commitment to mathematical truth. They disclose AI assistance with drafting the essay.
The 139-comment thread on Hacker News presses on the gap between an individual’s purpose and a profession’s survival.
What the thread adds
- GMoromisato — separates enjoyment, ability and paid work: “Sure, mathematics will survive and even prosper with AI, but what about the mathematicians?” klhs raises the related concern of attracting and supporting the next generation.
- dualvariable — proposes rewarding the person who makes a result understandable: “The first person to package it up for human consumption and get everyone to agree it is correct should get all the credit now.”
- abstractbill — supports the essay’s outlook: “the more of a Platonist I become, the less I care about who made a particular discovery.” Platonism treats mathematical truth as existing independently of its discoverer.
- jacquesm — asks why companies prize being first while researchers are urged toward detachment: “If ego wasn’t a factor, how come the companies that dump these results in the math community attach so much value to being the first to solve particular problems?”
Handles are pseudonymous; HN publishes no per-comment scores, and its ranking determines the fetched order. This is a slice of the thread, not a consensus.