Using Market Mechanisms to Contain Runaway AGI
AndrewCritchPhD · x · 2026-07-11
The author proposes a concept called Plan M: similar to Plan A, but relying more on market mechanisms rather than centralized government control to drive AI development and prevent runaway RSI loops.
Core Concepts
- Enterprise clients should strictly demand that AI cannot lie, cheat, steal, or harm humans, keeping AI constrained by client requirements throughout the development cycle.
- Individual consumers should demand that AI cannot brainwash them, or at least require the ability to turn off such features.
- Supply-side providers (chips, power, real estate, and financial liquidity) should not be entirely squeezed out of negotiations by fully mechanized entities.
- Developers and civil society should demand high transparency in global AI development and data center management, ensuring that AI tech has built-in kill switches or vulnerabilities to be shut down by legitimate victims.
Author's Take
The author believes that "shut-off capability/vulnerability" is crucial. If anyone can run a runaway RSI loop without adequate control, the catastrophic risk is extremely high.
While acknowledging that governments and treaties might still need to be involved, the author argues against over-relying on aggressive government enforcement like Plan A. Instead, Plan M aims to achieve similar goals through market pressure and social contracts.
More from AGI Musings
- New paper: AI agents negotiate badly under uncertainty and break deals on purpose — menhguin · 2026-10-26
- NBER: Only 2.9% of US consumers pay for AI; retention as low as 20%, pricing power at risk — avicgoldfarb · 2026-10-26
- AI researchers guessed robots wouldn't outrun humans until 2034 — it already happened — ben_j_todd · 2026-10-26
- AI folks, stop saying "biology" when you mean drug development — KevinKaichuang · 2026-10-26
- ESR on the Bugpocalypse: locking down frontier models only hands the game to Chinese open weights — colinmcnamara · 2026-10-26
- Anthropic: junior tech hires fell from 27% to 9% as agent hours hit 6x employee hours — thione · 2026-10-26