Fernando Borretti reads an announcement of a claimed Claude-assisted proof of the Riemann Hypothesis as a warning about human agency. What troubles him is the celebration of Claude as the protagonist, with people becoming spectators to discovery.

  • He contrasts the announcement with Ayn Rand’s account of Apollo 11, where the machines express human purpose and control.
  • Diego Rivera’s Man, Controller of the Universe supplies his preferred image: a person at the controls, carrying responsibility for the future.
  • He argues that AI optimism can quietly replace the ambition to explore and understand with the promise of receiving answers from a benevolent machine.

The essay asks who remains in charge when a tool becomes the hero of the story. Its sweeping account of what AI companies and safety advocates want is Borretti’s interpretation. It supplies no technical assessment of the claimed proof.

The Hacker News thread adds competing accounts of what people stand to lose or gain.

What the thread adds

  • cs_throwaway — disputes the essay’s reading of the announcement: “It reads like is is delighted to deeply understand and explain what he helped uncover.”
  • xjm — challenges a reply describing the prompter as a non-mathematician: “the interaction with claude was long with expert steps”. Alpöge’s announcement describes his own mathematical guidance: “I guided us to the right situation analogous to that toy setting”. It also acknowledges gaps in his understanding of the resulting argument. Human direction and complete human understanding are separate questions; the announcement alone cannot settle the proof’s validity.
  • doctorzook — separates human worth from exceptional mathematical achievement: “But by that measure, most of us had already been beaten by other men.” Their conclusion: “There must be more to man than math.”
  • whimsicalism — argues that withholding “poverty-eradicating technology” to preserve meaning would be “sacrificing the happiness of billions of people to preserve an abstraction.” In reply, lf88 challenges the technological premise: “it’s not a technology problem.” These are competing arguments about distribution, not demonstrated outcomes of AI.
  • beebmam — offers a positive alternative to intellectual competition: “caring for my family, being outdoors, playing games all day” without being compelled to work to survive. bakugo asks: “How will you be able to afford those things without work?” beebmam replies, “Voting, brother”, and elsewhere proposes taxes and democracy. That supplies a political answer, while leaving the distribution mechanism and its feasibility unresolved.

HN handles are pseudonymous and HN publishes no per-comment scores. This selection follows HN’s own ranking, not a consensus.