July 2, 2026
Proof? Receipts? Chaos.
The Fall of the Theorem Economy
A mathematician says papers are overrated, and the comments instantly turned feral
TLDR: A mathematician argued that the hardest part of math is finding the right idea, not writing a polished paper, and admitted one of his best results never got fully published. Commenters turned that into a bigger fight about AI, academic gatekeeping, and why math can feel brilliant but weirdly impossible to access.
A mathematician basically walked into the internet and said the quiet part out loud: the real work isn’t grinding out a formal proof, it’s having the big idea in the first place. In his essay, he admits one of his favorite results was never properly published, partly because he left academia for a machine-learning startup and partly because writing everything up felt like a soul-crushing chore. His argument is simple enough for non-specialists: in math, understanding matters more than stacking up official theorem trophies.
And wow, the commenters had feelings. One camp treated the piece like a mic drop, praising it as deeply wise and written by someone with towering expertise. Another camp immediately smelled a bigger cultural panic: is this really about math, or is it secretly about AI eating the prestige economy? One commenter cracked, “I see the AI panic has reached mathematics…,” which pretty much set the mood. Suddenly the thread wasn’t just about one unfinished proof — it was about whether human-friendly math is being replaced by machine-friendly “mathslop,” and whether future math might be readable only by bots and a tiny priesthood of experts.
Then came the classic internet subplot: math has a communication problem. One frustrated commenter agreed with the essay’s core point but blasted the field as hopelessly inward-looking, basically saying mathematicians love understanding while making very little effort to be understood. Add in disbelief over a paper apparently taking seven years from submission to acceptance, and the comment section turned into a full-blown referendum on prestige, publishing, AI anxiety, and whether modern math is genius at work or just elite homework with worse public relations.
Key Points
- •The author says his best theorem emerged while preparing for a conference talk in Lausanne, but he only mentioned it informally on a slide and never published it.
- •He attributes the lack of publication partly to having left academia and founded a machine-learning startup, leaving him without time to prepare a full proof.
- •The article presents Theorem 0.5 from an unpublished preprint on Garside categories as an example where proof was straightforward once the right framework and definitions were established.
- •The author argues that in mathematics, the hardest work can be building the conceptual language and identifying the right structure, rather than proving the theorem itself.
- •He states that his preprint remained available on arXiv, was cited dozens of times, and that its definitions were incorporated into later work, including a 700-page book on Garside theory.