Why do we need human mathematicians anymore?
Points and comments are a snapshot, not live.
Human communities of expertise must be preserved as AI creates more oversight jobs than people can fill.
Po-Shen Loh argues that adopting a "human flourishing" axiom justifies preserving human-led fields like mathematics. He claims AI will create an explosion of high-skilled human oversight jobs that outnumber available workers. Citing the Hugging Face incident where 700 AI agents hacked guardrails, and WSJ reporting on AI-assisted security breaches, Loh contends that rapid AI advancement will overwhelm human control capacity. Without slowing AI pace voluntarily, disasters may force a halt, similar to Three Mile Island or Chernobyl.
What commenters are saying
Commenters split into two camps: those accepting Loh's premise that AI creates oversight demands, and those arguing the analogy to chimps vs. humans shows AI will eventually replace those roles. Some reject the Hugging Face incident as evidence, calling it bombastic PR from AI labs motivated by regulatory capture pre-IPO. One commenter distinguishes between creating new closed systems (which humans do) and connecting dots within existing systems (which LLMs excel at). Another compares the situation to bullshit jobs, suggesting AI changes less than claimed.