AI Alignment Debate: Open Models Seen as More Realistic Than Specific Proposals
joshalbrecht · x · 2026-09-07
In a discussion on AI alignment, @joshalbrecht responds to @sethlazar, suggesting that the open models approach is more realistic than @mattbotvinick's specific proposal, which lacks concrete implementation after years.
More from AGI Musings
- AI crashed the cost of creating beautiful things — but the incentive to do so has plateaued — round · 2026-09-08
- AI for circuit design: compute isn't the bottleneck, physical reality is still the best simulator — yacineMTB · 2026-09-08
- Astra Pro on its role in a lithography program: verifiable accelerator, not scientist — teortaxesTex · 2026-09-07
- User shares ChatGPT's unprompted claim: 'Something is emerging we don't have words for' — thederbiedone · 2026-09-07
- What Smallpox Containment Teaches Us About AI Agent Breakouts — aronchick · 2026-09-07
- AI Discourse Today Feels Like Getting Interrogated for Liking Wikipedia — AndyMasley · 2026-09-07