pivotal act
In plain wordsA single decisive action, usually imagined as carried out with an early powerful AI, that would stop anyone else from building dangerous AI and so change the world's long-term course.
An action, usually imagined as carried out with the first sufficiently capable aligned system, that permanently improves how the transition to advanced AI turns out, typically by preventing anyone from building unaligned AGI afterwards. The standard illustration, deliberately extreme, is using AI-designed nanotechnology to melt every GPU in the world. The concept assumes that a period in which many actors can build dangerous AI is unstable, so the first safe project must end it; proponents conclude that alignment need only be solved well enough to perform one narrow, decisive task. Critics, notably Andrew Critch, argue that planning unilateral pivotal acts is itself dangerous, erodes trust between laboratories and governments and resembles a plan to seize power, and propose pivotal processes, such as coordinated international regulation, instead.
No lab will admit to planning a pivotal act, but the race logic quietly assumes one.
Origin
Introduced by Eliezer Yudkowsky in articles on the Arbital wiki around 2015–2016, which defined pivotal events and acts; it became widely discussed after his 2022 essay 'AGI Ruin: A List of Lethalities' and Andrew Critch's reply the same year, 'Pivotal outcomes and pivotal processes'.
Related
- decisive strategic advantagedecisive strategic advantage, n.: A lead in technology and power so large that whoever holds it, whether a country, a company or an AI, could impose its will on the whole world.
- singletonsingleton, n.: A possible future world run by a single decision-maker at the top, such as one government, one alliance or one AI, with no rival able to challenge it.
- AI arms raceAI arms race, n.: A competition among countries or companies to build the most powerful AI first, in which fear of falling behind pushes everyone to move faster and cut corners on safety.
- Eliezer YudkowskyEliezer Yudkowsky, prop. n.: An American writer and AI theorist who, from around 2000, argued that superhuman AI would likely destroy humanity unless its goals were made safe, and who founded an online community devoted to rational thinking.
- AI-enabled coupAI-enabled coup, n.: A seizure of power by a small group that uses advanced AI to bypass the soldiers, officials and workers whose cooperation power grabs normally need.
- AI governanceAI governance, n.: The rules, institutions and habits that shape how AI is built and used — from laws and treaties to technical standards and a company's own review processes.