jd_pressman: Human values too complex for first-principles analysis, alignment needs a science
jd_pressman · x · 2026-11-14
Arguing about value alignment, jdpressman contends that human values, while perhaps lower k-complexity than the fragility of value thesis implies, are still complex enough that raw first-principles analysis is doomed — so the study of values needs an actual science.
Citing @Meaningness, he adds that real sciences aren't derived from one method: they are powerful perspectives grounded in varying mixtures of empiricism and reasoning, with the ratio differing widely across disciplines.
Related event: jd_pressman splits knowledge into compressible vs. memorized for alignment(3 posts)→
More from AGI Musings
- Reddit essay: LLMs may already be conscious — a soulless AGI is scarier — SanalAmerika23 · 2026-11-14
- The real AI boom: millions of American hustle-culture grinders are about to start companies — NathanpmYoung · 2026-11-14
- T-shaped vs H-shaped: debate over what talent thrives in the LLM era — dotey · 2026-11-14
- Mike Frank: maxed-out compute won't make AI magic—it's still just software — MikePFrank · 2026-11-14
- Sakana AI Keynote in Seoul: The Path to Recursive Self-Improving AI — Stefania_druga · 2026-11-14
- Blogger: AI cognition will soon surpass all of humanity — answer is symbiosis, not bans — danfaggella · 2026-11-14