Do you want the US to "win" AI?

· hn top ai culture systems · Source ↗

TLDR

  • The author rejects US AI dominance framing and argues real progress means AI in hard possession by everyone, not a revocable API privilege.

Key Takeaways

  • Author self-identifies as someone who should embrace “neofeudalism” but rejects engineer-strongman society regardless of personal position in it.
  • Altman framed as product-focused; Musk criticized for zero serious open source institutions and no coherent societal role for normal people.
  • Anthropic’s “Mythos” fear campaign repeats the 2019 GPT-2 XL scare – same people who spun out of OpenAI in 2021 running the same playbook.
  • Good-actor test: who released AI to the world in hard possession vs. who controls it as a revocable API privilege.
  • SpaceX buying “AI bubble crap” dilutes its Mars mission; author argues Mars justifies a hard society, Earth does not.

Hacker News Comment Review

  • Commenters broadly reject the nation-state win framing, preferring diffuse capabilities and open source as the real win condition over US dominance.
  • US AI investment chart flagged as alarming; one commenter called the entire AI push “the largest scam in history” with no guaranteed payoff.

Notable Comments

  • @hiddencost: Flags the contradiction of calling EA evil while praising Altman, who agreed to supply the Pentagon with autonomous weapons.
  • @mindcrime: Frames the goal as “Open Source to win AI” – capabilities should remain evenly distributed, not concentrated in any country or org.

Original | Discuss on HN