Safety, ethics and risk

Effective accelerationism

In short

Effective accelerationism, or e/acc, is an online movement, mainly in Silicon Valley, that argues technology and markets, especially AI, should be pushed forward as fast as possible and opposes efforts to slow AI down for safety reasons.

Also called
e/acc, Accelerationism (AI)
Topic
Safety, ethics and risk
Level
Some background helps
First described
Set out in 2022 by pseudonymous X users including @BasedBeffJezos and @bayeslord, in a founding post and in "Notes on e/acc principles and tenets" (July 2022)12.

The movement began in 2022 among pseudonymous accounts on X (then Twitter). Its founding post, credited to @BasedBeffJezos, @bayeslord, @zestular and @creatine_cycle, describes e/acc as ideas that "seek to maximize the probability of the technocapital singularity"2. Notes published in July 2022 ground it in physics, arguing that the universe favours systems that capture more energy, and that markets and competition beat top-down control1.

Its supporters say slowing AI carries its own risks. The founding post argues that attempting to stall progress "isn't risk free"2. Venture capitalist Marc Andreessen's "Techno-Optimist Manifesto" of October 2023 says "any deceleration of AI will cost lives", names the precautionary principle as an "enemy", and lists @BasedBeffJezos among its "patron saints"3.

Fortune reports that Andreessen and Y Combinator chief executive Garry Tan added the e/acc label to their X profiles, and that members often criticise "decels", people they see as holding back progress, especially those wanting to slow AI for safety4. In December 2023 Forbes identified the person behind @BasedBeffJezos as Guillaume Verdon, a former Google quantum computing engineer and founder of the start-up Extropic4.

Critics argue the movement's tone works against its own aims. Fortune quotes New York Times columnist Ezra Klein, who wrote that treating so much of society "with such withering contempt will not speed up a better future"4.

Example

When debates break out over whether to pause training of more powerful AI models, e/acc supporters typically argue against any pause, while AI safety campaigners argue for one.

Why it matters

e/acc has influential backers in tech investment, so it shapes how some of the people funding AI think about safety rules and regulation.

Questions people ask

What does e/acc stand for?

Effective accelerationism, the name its founders gave the movement in 2022. Its founding post defines it as ideas that seek to maximise the probability of a "technocapital singularity"2.

Who is Beff Jezos?

The X handle of Guillaume Verdon, a former Google quantum computing engineer who founded the AI hardware start-up Extropic. Forbes revealed his identity in December 2023, and he has since spoken publicly under his own name4.

Is e/acc the opposite of AI safety?

Supporters say they want a better future, and argue that slowing technology has costs of its own23. But the movement does oppose many proposals from AI safety advocates, such as slowing AI development, and calls those who back them "decels"4.

Related terms

Sources

  1. Beff Jezos and bayeslord (2022), Notes on e/acc principles and tenets, Substack
  2. Effective Accelerationism (e/acc), verbatim repost of the inaugural e/acc post, e/acc newsletter
  3. Marc Andreessen (2023), The Techno-Optimist Manifesto, a16z
  4. Fortune (2023), "Beff Jezos" on billionaires and space (profile of Guillaume Verdon and e/acc)

Back to the glossary