AI capability growth as a tug of war: recursive self-improvement vs diminishing returns
CFGeek · x · 2026-11-08
CFGeek highlights Adam Marblestone's Claude-assisted growth model: AI capability growth is a tug of war between "AI improving AI" pushing toward hyperbolic growth and "diminishing returns" pulling toward a fizzle. The model splits progress into effective compute E = hardware H × software efficiency S, with dS/dt = a·R^λ·S^φ capturing parallelization penalty (λ<1) and ideas-getting-harder (φ<1), plus automation feedback tying research effort R to AI capability itself.
More from AGI Musings
- Rutgers mathematician: AI acknowledgements are a fad — judge all writing as AI-generated — AlexKontorovich · 2026-11-08
- tszzl: Comparing AI regulation to nuclear control is nonsense — models replicate infinitely — tszzl · 2026-11-08
- BitTorrent's Bram Cohen to mathematicians: your turn to debug AI proofs — hargup13 · 2026-11-08
- "Less worried about AI passing the Turing test than humans failing it" — TinfoilTricorn · 2026-11-08
- Scientists push back on Demis Hassabis' claim AI will cure all diseases in 10 years — coherence · 2026-11-08
- Stanford prof: AIxBio is enriched with grifters, integrity badly lacking — anshulkundaje · 2026-11-08