“Software brain” – the habit of fitting everything into controllable databases – explains the growing gap between tech industry AI enthusiasm and rising public hostility, not bad marketing.
Key Takeaways
Quinnipiac: majority of Americans think AI does more harm than good; only 35% excited; Gen Z anger rose from 22% to 31% YoY.
OpenAI spent $200M on TBPN podcast as a marketing fix; with 900M weekly ChatGPT users, people react to daily experience, not messaging.
DOGE’s database takeover failed because databases don’t match reality; software brain assumes the world conforms to the schema, not the reverse.
Law and code share formal structure and precedent reliance, but law is non-deterministic; AI legal tools built on software brain ignore this gap.
Amodei warned AI will replace entry-level white-collar work; Nadella says industry must earn social permission for energy use – neither has delivered either yet.
Hacker News Comment Review
Commenters note people clearly do want automation – dishwashers, 90M ChatGPT users – but demand the output be reliable enough to not require re-checking the work.
Several pushed back on conflating all “automation” with recent AI grievances; the article loses precision treating decades of automation history as one phenomenon.
One commenter argued people care about second-order effects like cheaper prices and flexibility, not the technology itself – a partial challenge to the “software brain gap” framing.
