OpenAI Says It Solved a Math Problem — Experts Say It's Garbage

OpenAI recently announced that its AI had solved a major problem in mathematics: proving that the Partition Principle does not imply the Axiom of Choice. This is a big deal in set theory, a field that studies the foundations of math. But when mathematicians actually looked at the paper, they were shocked. The paper is poorly organized, uses strange terminology, and cites unpublished lecture notes instead of proper sources. One expert called it "muddled" and said it would be immediately rejected by any serious journal.

The problem isn't that AI can't do math. It's that OpenAI dumped a confusing, low-quality paper and expected the math community to drop everything and verify it. As one mathematician put it, "When am I supposed to sift through a badly written paper?" Researchers are busy teaching, supervising students, and doing their own work. They don't have time to clean up after AI. This isn't a breakthrough; it's a burden.

This matters to everyone, not just mathematicians. AI is increasingly used in science, medicine, and law. If companies can claim "progress" with poorly documented results, it erodes trust. We need clear standards for AI-generated research. Should authors disclose AI use? Should chat logs be public? Without rules, we risk a flood of unreliable claims. OpenAI's stunt is a warning: just because AI can produce something doesn't mean it's useful or true. We all need to demand better.