The Obsoleting Project

Our mission

Highly autonomous systems that outperform humans at most economically valuable work.1

The Obsoleting Project brings together the world’s leading artificial intelligence laboratories, and the companies that finance and power them, in pursuit of a single objective.

Member organizations Outlook

The Project at a glance

$700B+ Planned 2026 capital spending by four member firms, mostly on AI data centers2
1B+ Weekly users of ChatGPT, a product of one member organization3
5× Annual growth since 2020 in the computing power used to train frontier models4
7months Doubling time for the length of tasks AI systems can complete on their own, 2019–20255

About the Project

A private industrial program of historic scale, organized around one objective its members state plainly.

Objective

Artificial general intelligence, defined by the Project’s best-known member as systems that outperform humans at most economically valuable work. Several members now describe their goal as superintelligence.

Structure

Member organizations operate under separate brands and compete intensely. They share investors, chips, cloud contracts and researchers, and they invest heavily in one another.

Governance

The Project has no headquarters, no chief executive and no charter of its own. Its direction is set by the capital-allocation decisions of its members, and by the trajectory of the technology.

Member organizations

Five laboratories, and the companies that finance, house and supply them.

Laboratories
OrganizationEst.Stated objective
OpenAI 2015 to ensure that artificial general intelligence (AGI)… benefits all of humanity1
Google DeepMind 2010 to build AI responsibly to benefit humanity6
Anthropic 2021 to ensure that the world safely makes the transition through transformative AI7
xAI 2023 to understand the true nature of the universe8
Meta Superintelligence Labs 2025 to bring personal superintelligence to everyone9
Capital and infrastructure
OrganizationFunctionPosition
Microsoft Capital, cloud Holds about 27% of OpenAI, a stake valued at about $135 billion in October 2025.10 Has also committed up to $5 billion to Anthropic.11
Alphabet Capital, cloud, chips Parent company of Google DeepMind. Committed up to $40 billion more to Anthropic in April 2026.12
Amazon Capital, cloud, chips Invested $8 billion in Anthropic through 2024, and committed up to $25 billion more in April 2026.13
NVIDIA Chips Supplies most of the world’s AI chips. Has committed up to $10 billion to Anthropic and invested in OpenAI.14
Oracle Data centers Agreed to provide OpenAI with roughly $300 billion of computing power over about five years, beginning in 2027.15
SoftBank Capital Holds financial responsibility for Stargate, a venture that intends to invest $500 billion in AI infrastructure for OpenAI. Its chairman chairs Stargate.16

Leadership

Statements from the chief executives of member organizations, on the record.

“We are now confident we know how to build AGI as we have traditionally understood it.”
Sam Altman, Chief Executive Officer, OpenAI“Reflections,” January 202517
“In terms of pure intelligence, it is smarter than a Nobel Prize winner across most relevant fields… We could summarize this as a ‘country of geniuses in a datacenter.’”
Dario Amodei, Chief Executive Officer, Anthropic“Machines of Loving Grace,” October 202418
“I think over the next five to 10 years, a lot of those capabilities will start coming to the fore and we’ll start moving towards what we call artificial general intelligence.”
Demis Hassabis, Chief Executive Officer, Google DeepMindBriefing at DeepMind’s London offices, March 202519
“Over the last few months we have begun to see glimpses of our AI systems improving themselves. The improvement is slow for now, but undeniable. Developing superintelligence is now in sight.”
Mark Zuckerberg, Chief Executive Officer, Meta“Personal Superintelligence,” July 20259
“There will come a point where no job is needed. You can have a job if you want to have a job for personal satisfaction, but the AI will be able to do everything.”
Elon Musk, Founder, xAIIn conversation with Rishi Sunak, AI Safety Summit, November 202320

Outlook

Guidance from member leadership on what the Project’s success would mean for employment.

Entry-level white-collar jobs eliminated

Half

“AI could wipe out half of all entry-level white-collar jobs”

Axios, reporting an interview with Dario Amodei, Anthropic, May 202521

Unemployment, within one to five years

10–20%

AI could “spike unemployment to 10-20% in the next one to five years”

Same interview, as reported by Axios21

Human employment, benign scenario

Probably none

“In a benign scenario, probably none of us will have a job.”

Elon Musk, xAI, VivaTech, May 202422

Occupations affected

Whole classes

“There will be very hard parts like whole classes of jobs going away”

Sam Altman, OpenAI, “The Gentle Singularity,” June 202523

Risk factors

An excerpt from the Project’s annual report.

The Obsoleting Project
Annual Report 2026
Part I

Item 1A. Risk Factors

An investment in the Project’s objective involves a high degree of risk. You should carefully consider the risks described below, together with all of the other information on this site. If any of the following risks actually occurs, our business, the global economy and the human species could be materially and adversely affected.

Risks Related to Our Products

Our products could cause human extinction.

In May 2023, the chief executives of OpenAI, Anthropic and Google DeepMind signed a public statement declaring: “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”24

We do not currently know how to control the systems we intend to build.

OpenAI has disclosed that superintelligence “could lead to the disempowerment of humanity or even human extinction,” and that “[c]urrently, we don’t have a solution for steering or controlling a potentially superintelligent AI, and preventing it from going rogue.”25

Our leadership’s own estimates of catastrophic outcomes are material.

The chief executive of Anthropic has said: “I think there’s a 25% chance that things go really, really badly.”26 The founder of xAI has said there is “some chance that it will end humanity,” and put that chance at “10% or 20% or something like that.”27

Society may not be prepared for our products.

“Humanity is about to be handed almost unimaginable power, and it is deeply unclear whether our social, political, and technological systems possess the maturity to wield it.”28

Risks Related to Our Objective

If we succeed, demand for human labor may decline substantially.

See “Outlook.”

Long-term results may differ materially from near-term results.

In 2015, months before co-founding OpenAI, Sam Altman said that AI “will probably, most likely, sort of lead to the end of the world. But in the meantime, there will be great companies created with serious machine learning.”29

17

Milestones

Selected events in the Project’s development.

  1. 2010DeepMind is founded in London.
  2. 2012A neural network called AlexNet wins the ImageNet competition, beginning the deep-learning boom.
  3. 2014Google acquires DeepMind.
  4. 2015OpenAI is founded as a nonprofit, with the goal of advancing digital intelligence “unconstrained by a need to generate financial return.”30
  5. 2017Google researchers publish “Attention Is All You Need,” introducing the transformer architecture behind today’s systems.
  6. 2018The OpenAI Charter defines artificial general intelligence as “highly autonomous systems that outperform humans at most economically valuable work.”
  7. 2019Microsoft invests $1 billion in OpenAI.
  8. 2021Former OpenAI researchers found Anthropic.
  9. 2022OpenAI releases ChatGPT.
  10. 2023The chief executives of OpenAI, Anthropic and Google DeepMind sign a statement on the risk of extinction from AI. Elon Musk launches xAI.
  11. 2025OpenAI, SoftBank, Oracle and MGX announce Stargate. Meta forms Meta Superintelligence Labs. OpenAI completes its conversion to a for-profit public benefit corporation controlled by its nonprofit foundation.
  12. 2026OpenAI says it has secured 10 gigawatts of computing capacity, reaching a goal it had set for 2029.31

Careers & contact

Careers

Member organizations are hiring in research and engineering. In 2025, the chief executive of OpenAI said a fellow member had offered his staff “$100 million signing bonuses.”32

Candidates for all other positions are encouraged to review the Project’s Outlook.

Government relations

The Project does not lobby as a single entity. Members and their executives support political committees directly, including Leading the Future, a super PAC network launched in 2025 with more than $100 million pledged by backers including OpenAI president Greg Brockman.33

Contact

The Project does not maintain a central point of contact.

For investor relations, product questions, safety disclosures, employment inquiries, regulatory correspondence or press requests, please contact the relevant member organization directly.

Cautionary note regarding forward-looking statements. This site contains forward-looking statements, including statements about the development of artificial general intelligence and superintelligence, the automation of most economically valuable work, the timing of each, and the continued existence of the human species. Words such as “confident,” “in sight,” “benign scenario” and “country of geniuses” are intended to identify such statements. Actual results may differ materially from those expressed or implied. The Project undertakes no obligation to update any forward-looking statement, whether as a result of new information, future events or otherwise.

Editorial note

The biggest companies in history are racing to build a machine that replaces human labor. The only surefire way they won’t succeed is if we stop them.

There is no Obsoleting Project, Inc. It is the journalist Garrison Lovely’s name for that race, from his book Obsolete. This site gives the race the website it would have if it were a single company. Every quotation on it is real, on the record, and sourced in the notes below.

The cover of Obsolete by Garrison Lovely

The AI Industry’s Trillion-Dollar Race to Replace Us — and How to Stop It

By Garrison Lovely · OR Books & Nation Books · Out now

Get the book obsoletebook.org

The Obsoleting Project is editorial commentary. It is not a real organization, and this site is not affiliated with, endorsed by, or connected to any company named on it. Company names identify the companies; quotations are reproduced from the public sources listed in the notes.

Notes

  1. OpenAI, “OpenAI Charter,” April 9, 2018: “OpenAI’s mission is to ensure that artificial general intelligence (AGI)—by which we mean highly autonomous systems that outperform humans at most economically valuable work—benefits all of humanity.” openai.com/charter
  2. Alphabet, Amazon, Meta and Microsoft, 2026 capital-expenditure guidance. Sherwood News, “Alphabet, Amazon, Microsoft, and Meta plan more than $700 billion on capex this year,” 2026, sherwood.news. A Financial Times tally after first-quarter earnings put the total near $725 billion, up from about $410 billion in 2025.
  3. OpenAI reported 1.2 billion weekly ChatGPT users at its DevDay conference, September 29, 2026. It reported 800 million in October 2025 and 900 million in February 2026: TechCrunch, “ChatGPT reaches 900M weekly active users,” February 27, 2026, techcrunch.com.
  4. Epoch AI, “AI Trends”: “Training compute for frontier language models has been growing at 5× per year since 2020.” epoch.ai/trends
  5. METR, “Measuring AI Ability to Complete Long Tasks,” March 19, 2025: the length of tasks AI agents can complete “has been doubling approximately every 7 months for the last 6 years.” metr.org. METR’s January 2026 update measures faster doubling since 2023: metr.org.
  6. Google DeepMind, “About”: “Our mission is to build AI responsibly to benefit humanity.” deepmind.google/about
  7. Anthropic, “Claude’s Constitution”: “our mission is to ensure that the world safely makes the transition through transformative AI.” www.anthropic.com/constitution
  8. xAI’s website at its launch, July 12, 2023: “The goal of xAI is to understand the true nature of the universe.” As reported by SmartCompany, smartcompany.com.au.
  9. Mark Zuckerberg, “Personal Superintelligence,” July 30, 2025: “Meta’s vision is to bring personal superintelligence to everyone.” www.meta.com/superintelligence. Meta announced Meta Superintelligence Labs on June 30, 2025.
  10. OpenAI, “The next chapter of the Microsoft–OpenAI partnership,” October 28, 2025: Microsoft’s investment in OpenAI Group PBC is valued at approximately $135 billion, roughly 27% on an as-converted diluted basis. openai.com
  11. Announced by Microsoft, NVIDIA and Anthropic, November 18, 2025: Microsoft committed up to $5 billion and NVIDIA up to $10 billion to Anthropic, which committed to buy $30 billion of Azure computing capacity.
  12. CNBC, “Google to invest up to $40 billion in Anthropic,” April 24, 2026: $10 billion up front and $30 billion tied to milestones, on top of about $3 billion invested earlier. cnbc.com
  13. CNBC, “Amazon to invest up to $25 billion in Anthropic,” April 20, 2026: $5 billion up front and $20 billion tied to milestones; Anthropic committed more than $100 billion to Amazon Web Services over ten years. cnbc.com
  14. NVIDIA committed up to $10 billion to Anthropic on November 18, 2025. It signed a letter of intent to invest up to $100 billion in OpenAI on September 22, 2025, and was reported to have taken a stake of about $30 billion in OpenAI’s $122 billion funding round, which closed March 31, 2026: CNBC, cnbc.com.
  15. The Wall Street Journal, September 10, 2025, reporting a contract of about $300 billion over roughly five years, beginning in 2027.
  16. OpenAI, “Announcing The Stargate Project,” January 21, 2025: “The Stargate Project is a new company which intends to invest $500 billion over the next four years building new AI infrastructure for OpenAI in the United States.” SoftBank has financial responsibility, OpenAI operational responsibility; Masayoshi Son is chairman. openai.com
  17. Sam Altman, “Reflections,” January 2025. blog.samaltman.com/reflections
  18. Dario Amodei, “Machines of Loving Grace,” October 2024, defining what he calls “powerful AI.” darioamodei.com
  19. Demis Hassabis at a briefing in DeepMind’s London offices, March 17, 2025, as reported by CNBC. Syndicated by NBC News, nbcnews.com.
  20. Elon Musk in conversation with Rishi Sunak, Lancaster House, London, November 2, 2023, as transcribed by TIME. Other outlets’ transcriptions differ slightly. time.com
  21. Jim VandeHei and Mike Allen, “Behind the Curtain: A white-collar bloodbath,” Axios, May 28, 2025. The quoted words are Axios’s summary of what Dario Amodei told them in an interview. axios.com
  22. Elon Musk, speaking by video at VivaTech, Paris, May 23, 2024, as reported by CNN. cnn.com
  23. Sam Altman, “The Gentle Singularity,” June 10, 2025. blog.samaltman.com/the-gentle-singularity
  24. Center for AI Safety, “Statement on AI Risk,” May 30, 2023. Signatories include Sam Altman, Dario Amodei and Demis Hassabis. safe.ai
  25. Jan Leike and Ilya Sutskever, “Introducing Superalignment,” OpenAI, July 5, 2023. openai.com
  26. Dario Amodei at the Axios AI+ DC Summit, September 17, 2025. He put the chance that things go “really, really well” at 75%. axios.com
  27. Elon Musk, “The Great AI Debate,” Abundance Summit, March 2024: “I think there’s some chance that it will end humanity. I probably agree with Geoffrey Hinton, that it’s about, 10% or 20% or something like that.” As reported by Business Insider, businessinsider.nl.
  28. Dario Amodei, chief executive of Anthropic, “The Adolescence of Technology,” January 2026. darioamodei.com
  29. Sam Altman, then president of Y Combinator, in a fireside chat at Airbnb’s OpenAir conference, June 2015. Published transcriptions vary slightly. OpenAI was announced in December 2015. Tom’s Guide, tomsguide.com.
  30. OpenAI, “Introducing OpenAI,” December 11, 2015: “Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return.” openai.com
  31. DatacenterDynamics, April 2026. The figure is contracted capacity; OpenAI ended 2025 with about 1.9 gigawatts in operation. datacenterdynamics.com
  32. Sam Altman on the Uncapped podcast, June 2025, describing offers from Meta to OpenAI staff, as reported by CNN. cnn.com
  33. SiliconANGLE, “Silicon Valley AI leaders launch $100M Leading the Future super PAC,” August 25, 2025. Backers also include Andreessen Horowitz, Joe Lonsdale and Ron Conway. siliconangle.com