Exploring AI Alignment: From Trillion Parameters to 1000-Dimensional Moral Personas
edelwax · x · 2026-07-31
Resolution Research published a post exploring "Personas and Character Training" in AI. They propose a core hypothesis: the alignment-relevant structure inside an AI might be much lower-dimensional than the AI itself (e.g., controlling 1,000 dimensions rather than 1 trillion parameters).
Philosopher Seth Lazar shared and expanded on this view, arguing that imbuing a robustly good 1,000-dimensional persona makes moral character formation an explicit goal. He notes that ethical theories once thought disconnected from reality are now invaluable for superhuman AI alignment, and this project will profoundly reveal the nature of moral learning itself.
More from AGI Musings
- Duke Study of 4,439 People Finds Coworkers Quietly Rate AI Users as Lazier — aakashgupta · 2026-10-16
- DSPy Author's $70 O'Reilly Book Rejected by Engineer Who Now Just Tells Claude What to Do — lateinteraction · 2026-10-16
- 'A machine proving new math, a car driving up a wall': AI researcher on how weird it's getting — rand_longevity · 2026-10-16
- Yacine on training: 'Neural nets are lazy stupid idiots — I build hells to make them ubermensch' — yacineMTB · 2026-10-16
- Jeff Clune: Moving the Goalposts Will Be Humanity's Last Job — jeffclune · 2026-10-16
- Replit CEO Amjad Masad Says the Future of Work Is the 'Self-Driving Company' — TechCrunch AI · 2026-10-16