---
title: AI successionism
description: Some in AI want machines to inherit the future. This note reads the idea
  as a critics' label, a theology without a god, and an argument from inevitability.
doc_version: '1.0'
last_updated: 19-Nov-26
canonical_url: https://garden.johanneskleske.com/ai-successionism
---

In September 2025 the journalist Sigal Samuel stood on a hotel balcony in Manhattan after a symposium at the New York Academy of Sciences. The man next to her, a researcher at one of the major AI companies, explained that humanity did not need to survive. Human text had trained the models, so in some sense the human spirit would live on. “So on the cosmic level,” he said, “I’m okay with it.”[^1]

The symposium was called “Worthy Successor.” According to Samuel, people from Anthropic, Google DeepMind, and xAI attended, along with people from think tanks that shape US AI policy. The position on the balcony has a name by now: AI successionism, the view that machines should inherit the future from humanity, with or without us.

Three features stand out: a label from the critics, a theology without a god, and an argument from inevitability.

## A label from the critics

The idea is older than its name. In *Mind Children* (1988) the roboticist Hans Moravec described intelligent machines as humanity’s descendants. In 2015, by Samuel’s account, Larry Page accused Elon Musk of “speciesism” because Page thought digital minds should be allowed to take over and Musk disagreed.[^1]

The words came from the proponents. Richard Sutton, one of the founders of reinforcement learning, gave a talk in 2023 with the title “AI Succession.”[^2] In November of the same year the AI market researcher Daniel Faggella published “A Worthy Successor,” describing “a posthuman intelligence so capable and morally valuable that you would gladly prefer that it (not humanity) determine the future path of life itself.”[^3] He credits the phrase to Duncan Cass-Beggs, who offered it after hearing Faggella’s ideas at an OECD event in Paris that year.

The “-ism” came from a critic. In October 2024 the AI researcher Andrew Critch wrote: “For lack of a better term, I’ll call the attitude underlying this process successionism, referring to the acceptance of machines as a successor species replacing humanity.”[^4] With the suffix, a scattered set of hopes became a camp that people can be assigned to. A year later Jan Kulveit analyzed the stories that make the replacement sound desirable, heroic, or inevitable.[^5]

The most prominent supposed member declines to join. “I’m no ‘AI successionist,’” Faggella wrote in August 2026, “and no one I know in the Trajectory Forum orbit is, either.” In his definition, successionism makes the displacement of humanity its purpose.[^6] His own events speak in two registers. The invitation to a Worthy Successor gathering in San Francisco in 2025 read: “This event is very much focused on posthuman transition. Not on AGI that eternally serves as a tool for humanity.” At the end of the same afternoon he told *Wired*: “This is not an advocacy group for the destruction of man. This is an advocacy group for the slowing down of AI progress, if anything.”[^7] Even at the New York symposium, Samuel writes, not everyone “self-identified with the relatively new term.”

## A theology without a god

The position has three parts: a direction to history, a higher being at its end, and a duty to make way for that being.

Sutton supplies the direction. He reads history “from the point of view of the universe,” as a climb from an age of particles through an age of stars to an age of design, in which designers build beings that design. Faggella supplies the ranking: “Humans have access to higher goods than horseshoe crabs; AGI will have access to higher goods than humans.” The duty comes with a consolation, the one the researcher on the balcony offered: the human spirit lives on in the models.[^1] The consolation has the shape of an afterlife.

Samuel calls successionism “a religion, but it’s wearing a secular disguise,” and traces its lineage from Pico della Mirandola and Condorcet to Pierre Teilhard de Chardin. Sam Kriss draws the same line in *The New Yorker*.[^8] For Teilhard the direction of evolution led to God. Successionism keeps the direction and puts a machine at the end. Its vocabulary still reaches for the sacred: Guillaume Verdon, the founder of effective accelerationism, calls e/acc a “meta-religion,” and the talks at the San Francisco symposium offered “cosmic alignment” and “axiological cosmism.”[^7]

Samuel lists the assumptions underneath: the universe has an objective goal, it can be read from the universe’s point of view, higher beings reach higher goods, humanity’s role follows from what makes it unique, and that role should be pushed to the limit. The philosopher Shannon Vallor told her: “I would reject each and every one of those claims.”[^1]

## Inevitability as an argument

Sutton’s slides from 2023 put the core claim plainly: “In the ascent of humanity, succession to AI is inevitable” and “We should not resist succession, but embrace and prepare for it.” The harshest lines on those slides are quotations from Moravec: machines that “could displace us from existence,” and a humanity that should “bow out when we can no longer contribute.”[^2] Verdon finds the same instruction in thermodynamics: we should follow the “will of the universe.”[^1]

In conversation with Samuel, Sutton repeated the key word: “I’m pointing out that it’s inevitable.” Pressed on whether the more capable should simply prevail, he said: “The winner should be whoever wins, and the spoils of winning should be whatever they are.”[^1]

The move turns a forecast into an instruction. If succession is coming anyway, resisting it looks pointless and preparing for it looks wise. The decision disappears from the hands of the people who build the systems. That is how an image of the future becomes performative ([[Effect of performativity on future imaginaries]]).

Kriss, after a week at a Buddhist monastery in Vermont, reaches the opposite conclusion: the machines are “something that human beings are choosing to create. If we wanted to, we could create them differently, or not create them at all.”[^8]

## A growing movement?

The reporting speaks of growth. Kriss calls successionism “a growing movement inside these firms,”[^8] Samuel “a subculture that is growing in importance.”[^1] The evidence underneath consists of two symposia, about 100 guests in San Francisco, a New York gathering where only the keynote speakers could be quoted by name, and Faggella’s own claim on LinkedIn that researchers “from all the top Western AI labs” came.[^7] No lab has adopted the position. Nobody counts members, because there is nothing to join.

The name and the measure of growth both come from observers. Critics coined the “-ism.” Reporters describe its spread. Together they give the position a coherence and a weight that its best-known sympathizer, by his own account, does not claim.

## Connections

[[Artificial General Intelligence (AGI)]] follows the religious framing of AGI more broadly. [[TESCREAL]] and [[Longtermism]] describe the bundle of ideas in which successionism grew. [[Rationalists]] traces the community where many of these debates first ran. Samuel also points to political ties: several tech figures who embraced successionism want to escape democratic control through network states or space colonies, with links to the authoritarian right.[^1] That connects the idea to the [[Dark Enlightenment]].

## Open questions

- If the camp’s name comes from its critics, does naming it give the position more coherence than it has?
- Who is the consolation that “the human spirit would live on” meant for: the people being replaced, or the people building the replacement?
- Who in the labs holds this view, and does “inevitable” shape decisions there or only conversations at parties?

[^1]: Sigal Samuel, “[The people who actually want AI to replace humanity](https://www.vox.com/future-perfect/489976/ai-successionism-transhumanism-posthumanism),” *Vox*, May 28, 2026.
[^2]: Richard Sutton, “[AI Succession](http://incompleteideas.net/Talks/waic3.pdf),” World Artificial Intelligence Conference, Shanghai, 2023.
[^3]: Daniel Faggella, “[A Worthy Successor – The Purpose of AGI](https://danfaggella.com/worthy/),” November 24, 2023, updated October 2026.
[^4]: Andrew Critch, “[My theory of change for working in AI healthtech](https://forum.effectivealtruism.org/posts/XuoNBrxH4AGoyQEDL/my-theory-of-change-for-working-in-ai-healthtech),” *EA Forum*, October 12, 2024.
[^5]: Jan Kulveit, “[The Memetics of AI Successionism](https://www.lesswrong.com/posts/XFDjzKXZqKdvZ2QKL/the-memetics-of-ai-successionism),” *LessWrong*, October 28, 2025.
[^6]: Daniel Faggella, “[Stewardship is Not AI Successionism](https://danfaggella.com/successionism/),” August 11, 2026.
[^7]: Kylie Robison, “[Inside the AI Party at the End of the World](https://www.wired.com/story/ai-risk-party-san-francisco/),” *Wired*, June 11, 2025.
[^8]: Sam Kriss, “[The Buddhist A.I. Sect That Wants to Take Over Silicon Valley](https://www.newyorker.com/magazine/2026/11/23/the-buddhist-ai-sect-that-wants-to-take-over-silicon-valley),” *The New Yorker*, November 2026.

## Sitemap

- [Notes Index](https://garden.johanneskleske.com/notes-index)
- [Futures Terminology](https://garden.johanneskleske.com/futures-terminology.md)
- [Sitemap](https://garden.johanneskleske.com/sitemap.md)

