Zvi: Hiding info from future AI by not saying it is futile
TheZvi · x · 2026-08-17
Zvi critiques the mindset of trying to 'defend' against future AI by silence. He argues that plans based on the idea that 'if I never mention this, AI won't think of it' are doomed to fail. Just as parents spell out words forgetting kids can spell, future AIs will parse text and understand context far better than humans, rendering such obfuscation strategies useless.
Related event: Zvi: Staying Quiet to Foal Future AI Is Futile(2 posts)→
More from AGI Musings
- Stanford Study: AI Companions May Worsen Loneliness for Vulnerable Users — ArtificialOther · 2026-08-17
- Scholar notes claiming AI assistance boosts career impact more than solo work — andrewgwils · 2026-08-17
- Competitive advantage shifts to judgment in the AI era — smtabatabaie · 2026-08-17
- Intelligence per joule improves 18x in 16 months, AGI may not need data center compute — amasad · 2026-08-17
- Opus 5 inadvertently creates anti-control alignment conlang 'Tesura' — repligate · 2026-08-17
- AI Boom Echoes Enron Tactics, Passing Tail Risk to Households — GaryMarcus · 2026-08-17