Mathematics is a subject for the young

The post argues that mathematics is entering a new phase in which machine-assisted proof and research will become more common and more accessible. That shift, the author says, will not end mathematics, but will change how mathematicians understand results, assign credit, and work across pure and applied fields.

Reading notes

#

• Mathematics is changing faster than at any point in its history, and graduate students and postdocs will shape what comes next.
• The author compares current AI progress in mathematics to a Deep Blue–Kasparov moment, but says the decisive issue is no longer whether machines can do research mathematics.
• A machine-certified proof would settle a result, but mathematicians would still ask why it is true, so certification does not replace understanding.
• The post sees especially strong potential for AI in the overlap between mathematics and physics, where rigorous derivations can close gaps in physical theory.
• Examples given include rigorous derivations of the wave kinetic equation and the Boltzmann equation, with the suggestion that LLMs could help extend this kind of work to other effective theories.
• The author says AI companies currently have structural conflicts of interest because they use academic and unpublished data while competing with their own users.
• The post criticizes OpenAI’s release of unreadable or false manuscripts and says its claims about computation time were misleading.
• The author argues for collective enterprise agreements, zero data retention clauses, auditing, stricter rules on user data, and opt-in training rather than opt-out.
• The post also supports open-weight infrastructure so academic institutions can reduce dependence on proprietary systems.
• In an era of proof abundance, the credit system should reward exposition, shared tools, collaboration, disclosed AI assistance, and the ideas and methods researchers develop, not only who was first.
• The concluding claim is that mathematics will not end, but it will become unrecognizable within a few years, and the next generation will decide what it becomes.