Three prevalent themes in the discussion
- Human‑AI collaboration vs. full autonomy
Many commenters stress that effective use of LLMs in mathematics requires human steering, oversight, and understanding—not just letting the model run for long stretches. - Semkas: “I feel like waiting on the output of an LLM for 40 hours feels like it is completely antithetical to what makes classic Hackathons appealing / educative.”
- falcor84: “The best teams are those that coordinate with the AI agents in relatively quick cadence, generally giving it small tasks and steering it often.”
-
brian‑bfz: “Can we set high standards for LLM use? Can we highlight the roles of a mathematician beyond proof generation? … We’re very open to criticisms.”
-
Ethical and societal concerns about AI deployment
Several participants raise worries about the motives of big AI labs, environmental harms, job displacement, and the risk of turning mathematicians into cheap validators for corporate models. - youoy: “Because the companies that run frontier models are malevolent by every metric. They are destroying the environment … Seriously why would you use them? Your use only emboldens them; making you complicit in their nefarious success.”
- tzs: “Probably many mathematicians want answers to the questions … the big AI companies agreed to sponsor them to find out because it is good publicity for the companies.”
-
reasonableklout: “It’s a noble goal to change the incentives, but how will you prevent the headlines … being ‘students prove Collatz conjecture with Claude’ and instead be ‘students give great explanation …’?”
-
Emphasis on problem selection, understanding, and communication over raw proof generation
The value of the event is seen in choosing worthwhile problems, grasping the proofs, and being able to explain them—rather than merely producing AI‑generated output. - Aboutplants: “If the goal is to accomplish something then why limit yourself with available tools? Use all available tools to accomplish the goal/task, it’s where a lot of incredible learning came out of.”
- jhonof: “I think the purpose of an event like this would be to optimize the process so that it isn’t just occasionally asking an AI to keep going.”
- a2ff6eeb0: “The best progress will be made by a team that purely spends their time taking a list of open problems and promoting ‘solve
’, without actually trying to understand anything.” (presented as a cautionary extreme) - brian‑bfz: “Solving a problem is the easy part with AI. Selecting a problem worth solving is half of the challenge.”