Goertzel's new post: using d-calculus to find the beneficial-superintelligence attractor
burny_tech · x · 2026-10-06
Ben Goertzel's new Substack essay "Finding the Beneficial-Superintelligence Attractor" tackles a core question: as an AGI keeps rewriting its memory, reasoning habits, and even its goal-handling code, what stops it from shedding its original value commitments one reasonable-looking upgrade at a time?
- He applies his new d-calculus — math for minds that fork, merge, learn and rewrite themselves — to argue under what conditions self-improving systems tend to preserve goals and may evolve into the basin of the beneficial-superintelligence strange attractor.
- The post revisits classic late-90s debates on goal preservation under self-modification and wireheading.
- It's a follow-up to his previous long math post, with more diagrams and planned OmegaHive experiments.
More from AGI Musings
- Phase Lock manifesto argues brain-computer interfaces can help align AI and preserve human agency — EricMTrautmann · 2026-10-06
- LLM-assisted 22-page preprint settles Hilbert's 12th problem and Zauner's conjecture — basedjensen · 2026-10-06
- Nikita Bier: AI is capable enough — only non-coder adoption speed limits disruption — _AustinCalvert_ · 2026-10-06
- OpenAI reportedly spent tens of millions in compute to crack Navier-Stokes in days — haider1 · 2026-10-06
- Nearly half of DOL-recognized jobs have zero agentic AI tool coverage, Cohere finds — Cohere_Labs · 2026-10-06
- Goertzel's 2010 classic: 'The Singularity Institute's Scary Idea, and why I don't buy it' — burny_tech · 2026-10-06