Choose a letter · navigate the seven ideologies
TTranshumanismEExtropianismSSingularitarianismCCosmismRRationalismEAEffective AltruismLLongtermism
S Part of the TESCREAL bundle

Singularitarianism

Building God, on a deadline.

The belief that machine intelligence will soon surpass our own and trigger a runaway “intelligence explosion” — the Singularity — after which human history as we know it ends. The third letter of TESCREAL, and the one that turns the bundle into an engineering project.

Singularitarianism takes transhumanism's promise of transcendence and names the mechanism: a self-improving artificial intelligence that, once smarter than us, redesigns itself again and again until it leaves humanity behind in a matter of months or hours. For its prophets the Singularity is a technological Rapture to be welcomed and accelerated; for its doomsayers it is an extinction event to be prevented at any cost. Either way it casts a small priesthood of insiders as the people who decide the fate of all future life — and both wings, critics note, agree the stakes are infinite.

Overview

What it is

The point of no return

The Singularity is borrowed from physics: the point at a black hole's center where the known laws break down and prediction becomes impossible. Applied to technology, it names the moment when machine intelligence exceeds human intelligence and begins improving itself faster than we can follow. Beyond that horizon, Singularitarians argue, the future is literally unknowable, because it will be shaped by minds we cannot comprehend.

The mechanism is recursive self-improvement. A machine smart enough to do AI research can build a smarter machine, which builds a smarter one still — a feedback loop that, once it starts, runs away from human control. The mathematician I. J. Good called the result an “intelligence explosion,” and observed that the first ultraintelligent machine would be “the last invention that man need ever make.”

Rapture for the nerds

The concept was formalized in 1993 by mathematician and novelist Vernor Vinge in a paper delivered at a NASA symposium: “Within thirty years, we will have the technological means to create superhuman intelligence. Shortly after, the human era will be ended.” But it was Ray Kurzweil who turned a hypothesis into a worldview — and into something openly theological.

In The Singularity Is Near (2005) Kurzweil dated the event to 2045 and described it in the language of religion. Asked whether there is a God, he answered: “Not yet, but there will be.” Academics have catalogued the parallels: the scholar Egil Asprem reads Kurzweil's Singularitarianism as “contemporary esotericism,” a millenarian faith with the Singularity as its Rapture, mind-uploading as resurrection, and The Singularity Is Near as its scripture. Critics inside the movement nicknamed it “the Rapture for the nerds.”

The doom wing

Not every believer is celebrating. Where Kurzweil sees salvation, Eliezer Yudkowsky sees the default outcome as human extinction. His thesis: superintelligence is coming, almost every way of building it kills everyone, and the people who understand this carry a moral obligation to solve “alignment” above all else. To that end he founded the Singularity Institute — later MIRI — and spawned the rationalist community as a recruiting and training ground for the problem.

The two wings disagree about the ending but share the premise, and that premise is what critics find most consequential: that a tiny group of insiders in a handful of labs is steering an event of cosmic, infinite stakes. By 2025 Yudkowsky and Nate Soares were arguing the case in a trade book titled, without hedging, If Anyone Builds It, Everyone Dies — while the labs the movement helped seed, DeepMind and OpenAI, raced to build the very thing it warned about.

Core tenets

What its believers hold

The doctrines that recur across the movement's founders, theorists, and builders.

01

Intelligence is the master variable

Everything else — wealth, health, power, survival — flows downstream of intelligence. Whoever builds the smartest system controls the future.

02

Self-improvement runs away

An AI able to improve itself enters a feedback loop. Each gain accelerates the next, so progress goes from gradual to explosive — fast enough to escape human oversight.

03

The future is a wall we can't see past

After superintelligence arrives, prediction breaks down. Today's institutions, economics, and politics become irrelevant overnight. There is a clean break with all prior history.

04

The Singularity is near — and inevitable

Exponential trends in computing make it a question of when, not if. Kurzweil's “Law of Accelerating Returns” treats the date as something you can read off a graph.

05

It is the most important event ever

Getting the Singularity right (or wrong) outweighs every other human concern. This justifies treating AI as a near-religious priority — and its builders as people of world-historic responsibility.

06

A vanguard must steer it

Because the stakes are infinite and the public doesn't understand, an enlightened minority — researchers, founders, funders — has both the right and the duty to decide how it unfolds.

Key concepts

The vocabulary

The specialized terms you need to read the movement in its own words.

The Singularity
The hypothesized point at which machine intelligence surpasses human intelligence, after which technological change becomes so rapid and profound that the future is unknowable.
Intelligence explosion
I. J. Good's idea that a sufficiently smart machine could recursively design ever-smarter machines, leaving human intelligence far behind in a short burst.
Recursive self-improvement
An AI rewriting and upgrading its own code or successor designs, each version more capable than the last — the engine of the intelligence explosion.
Hard vs. soft takeoff
Whether superintelligence arrives in hours or days (hard / “FOOM”) or over years and decades (soft). The faster the takeoff, the less time to react or correct course.
AI alignment
The problem of making a superintelligent system reliably pursue human-intended goals. Yudkowsky's wing treats failure here as the most likely cause of human extinction.
Law of Accelerating Returns
Kurzweil's claim that information technologies improve exponentially, not linearly, making the Singularity's arrival predictable and (he argues) inevitable around 2045.
Manhattan Project for AI
The recurring framing of AGI as a wartime-scale crash program — invoked by builders like Demis Hassabis and, later, by US policymakers urging a national race to superintelligence.
AI 2027
A 2025 forecasting scenario by ex-OpenAI researcher Daniel Kokotajlo and others, modeling a near-term path to superhuman AI — armies of AI copies automating research within a few years.
Timeline

How it unfolded

1965
I. J. Good names the “intelligence explosion”
The cryptographer who worked with Turing at Bletchley Park writes that an ultraintelligent machine could design even better machines — “the last invention that man need ever make.” The seed of every takeoff argument.
1993
Vernor Vinge — “The Coming Technological Singularity”
At a NASA-sponsored symposium, Vinge gives the idea its modern name and a deadline: superhuman intelligence within thirty years, after which “the human era will be ended.”
2000
Yudkowsky founds the Singularity Institute
A teenage-prodigy autodidact who once predicted a 2020 Singularity founds the Singularity Institute for Artificial Intelligence — later renamed MIRI — to make superintelligence safe rather than merely fast.
2005
Kurzweil — “The Singularity Is Near”
Ray Kurzweil dates the Singularity to 2045 and frames it in theological terms. The book becomes the movement's scripture; “Not yet, but there will be [a God]” becomes its most quoted line.
2008
Singularity Summit & Singularity University
Thiel-funded Singularity Summits at Stanford turn the subculture into a scene; Kurzweil and Larry Page launch Singularity University at NASA Ames — critics call it the faith's seminary.
2010
DeepMind — and a Thiel cheque
Demis Hassabis co-founds DeepMind with the explicit aim of “solving intelligence.” Peter Thiel — introduced to the founders through the Singularitarian network — provides early backing. Google buys it in 2014.
2015
OpenAI founded
Sam Altman, who credits Yudkowsky as a critical influence on the field, co-founds OpenAI to build AGI “for all of humanity.” The doom wing's warnings and the builder wing's ambition fuse into one industry.
2025
“If Anyone Builds It, Everyone Dies” / AI 2027
Yudkowsky and Nate Soares publish a trade book arguing any superintelligence built with current methods kills everyone, while the “AI 2027” scenario models superhuman AI arriving within a couple of years. Prophecy meets countdown.
Key figures

The cast

Founders, theorists, builders, and financiers of the project.

Eliezer Yudkowsky
Doom prophet

Eliezer Yudkowsky

b. 1979

Founded the Singularity Institute / MIRI and the rationalist community. Argues misaligned superintelligence is the default road to extinction; co-author of “If Anyone Builds It, Everyone Dies.”

Wikipedia →
Ray Kurzweil
Prophet

Ray Kurzweil

b. 1948

Turned the Singularity into a worldview and dated it to 2045. “Not yet, but there will be [a God].” Co-founded Singularity University with Larry Page.

Wikipedia →
Peter Thiel
Patron

Peter Thiel

b. 1967

Funded the Singularity Institute and the Stanford Singularity Summits, and was an early backer of DeepMind. The movement's most consequential financier.

Wikipedia →
Demis Hassabis
Builder

Demis Hassabis

b. 1976

Co-founded DeepMind to “solve intelligence,” framing AGI as a Manhattan-Project-scale effort. Now leads Google's AI; shared the 2024 Nobel Prize in Chemistry.

Wikipedia →
Sam Altman
Builder

Sam Altman

b. 1985

OpenAI CEO; credits Yudkowsky as a critical influence on the field. Describes AGI as “magic intelligence in the sky” — the Singularity recast as product roadmap.

Wikipedia →
Nate Soares
Ideologue

Nate Soares

b. 1987

President of MIRI and co-author with Yudkowsky of “If Anyone Builds It, Everyone Dies” (2025) — the doom wing's argument made into a mass-market warning.

Wikipedia →
Key texts

The canon

The essays, declarations, and books that built the doctrine.

The Coming Technological Singularity

Vernor Vinge

The founding text, delivered at a NASA symposium. “Within thirty years… Shortly after, the human era will be ended.”

Read ↗
1993

The Law of Accelerating Returns

Ray Kurzweil

The essay that supplied the math: exponential, not linear, progress — making the Singularity, in Kurzweil's telling, inevitable.

Read ↗
2001

The Singularity Is Near

Ray Kurzweil

The movement's scripture. Dates the Singularity to 2045 and describes it in openly theological terms.

Read ↗
2005

Intelligence Explosion Microeconomics

Eliezer Yudkowsky

MIRI technical report arguing that recursive self-improvement could produce a fast, decisive takeoff — the case for hard FOOM.

Read ↗
2013

Superintelligence: Paths, Dangers, Strategies

Nick Bostrom

The academic statement of takeoff and the control problem that pulled Musk, Altman, and policymakers into the alignment debate.

Read ↗
2014

If Anyone Builds It, Everyone Dies

Eliezer Yudkowsky & Nate Soares

The doom wing's mass-market case: building superintelligence with today's methods means human extinction.

Read ↗
2025
In their own words

On the record

Within thirty years, we will have the technological means to create superhuman intelligence. Shortly after, the human era will be ended.
— Vernor Vinge, 1993
Is there a God? Not yet, but there will be.
— Ray Kurzweil
The first ultraintelligent machine is the last invention that man need ever make.
— I. J. Good, 1965
If anyone builds it, everyone dies.
— Eliezer Yudkowsky & Nate Soares, 2025
The critique

The case against

Singularitarianism presents itself as hard-nosed forecasting — just extrapolating the curves. Critics argue it is eschatology in a lab coat: a faith whose central event is always about twenty years away, whose failures to arrive are explained as delays, and whose true function is to hand a small priesthood of technologists authority over everyone else's future.

It's a religion in disguise

The structure maps exactly onto apocalyptic Christianity — Rapture, resurrection, original sin, a saved remnant. Scholars like Egil Asprem classify Kurzweil's version as millenarian “contemporary esotericism,” not science.

Unfalsifiable prophecy

The date keeps moving — 2020, 2029, 2045 — and every miss is reframed as “not yet.” A claim that can never be wrong is not a prediction; it's an article of faith.

Doom and boom serve the same end

Whether the AI saves us or kills us, both wings agree it is the most important thing in history and that insiders must steer it. The framing concentrates power and capital in the very labs racing to build it.

Infinite stakes justify anything

If the Singularity is cosmic salvation or extinction, then present human concerns — jobs, rights, the eight billion alive now — shrink to a rounding error against an imagined future. That math has been used to excuse almost anything.

Inside the bundle

How it connects

Transhumanism does not stand alone. It feeds — and is fed by — the other letters of TESCREAL.

The network

The full web

People, organizations, and the ideologies this one bleeds into. Drag nodes, hover to isolate, scroll to zoom — or pop it out fullscreen.

drag · hover · scroll to zoom
Sources

Citations & further reading

Primary documents and reporting used on this page. See the full References library → for the complete source list.

  1. 01
    The Coming Technological Singularity (1993 NASA paper) ↗Vernor Vinge · SDSU
  2. 02
    Technological singularity ↗Wikipedia
  3. 03
    Singularitarianism ↗Wikipedia
  4. 04
    Intelligence explosion ↗Wikipedia
  5. 05
    The Singularity Is Near ↗Wikipedia
  6. 06
    Intelligence Explosion Microeconomics (Technical Report 2013-1) ↗Eliezer Yudkowsky · MIRI
  7. 07
    Singularity University ↗Wikipedia
  8. 08
    The Magus of Silicon Valley: Apocalypse and God-Making in Kurzweil's Transhumanism ↗Egil Asprem · DiVA Portal
  9. 09
    The TESCREAL Bundle: Eugenics and the Promise of Utopia through AGI ↗Gebru & Torres · First Monday (2024)
  10. 10
    AI 2027 (forecasting scenario) ↗Kokotajlo et al.
Singularitarianism — connection network