AI building AI · 1 Jan 2023 – 22 Nov 2026

How the story was told, week by week

Whether AI is speeding up the making of AI: what the press reports, and what the measurements show. 4,262 headlines from 1,487 news sites since 1 January 2023, 1,370 of them on one of the questions we follow. The river below is those 1,370, one colour per question. Where it swells, the world was talking about one thing; where a colour thins, a thread went quiet. Since 23 September 2026 our own stories ride on it: 4 so far, each linking to the articles behind it.

20 days since 1 January 2023 are not in the record yet; they are drawn as gaps, not as quiet.

The river headlines per week, one colour per question

The race to build superintelligence 207AI that improves itself 160When will AGI arrive? 116Stopping dangerous superintelligence 247Testing claims of AGI 157The business bet on AGI 163Countries preparing for AGI 135Chips and cloud for AGI 185

The chapters as the record wrote them

A chapter starts when a different question leads the week, or when a week is far louder than the chapter before it. Each card is the leading question's loudest week in the chapter, exactly as that question's log has it; the date opens the week there.

202626 chapters · 556 headlines
  1. "We won't allow SI to be stifled": Trump opposes global regulation of superintelligence at 33rd APEC summit

    16 Nov – 22 Nov · 5 headlines
  2. Anthropic says AI could be nearing recursive self-improvement

    2 Nov – 8 Nov · 8 headlines
  3. This Startup Is All In on Recursive Self-Improvement. Its Latest Model Is Scary Good

    19 Oct – 25 Oct · 2 headlines
  4. AI researchers warn companies rushing self-improving systems despite safety risks

    28 Sep – 4 Oct · 6 headlines
  5. Sanders and Casar propose ban on artificial superintelligence

    21 Sep – 27 Sep · 27 headlines · 2 of our stories
  6. 'Pause AI development, ban superintelligence' - Sanders to Trump, Xi

    7 Sep – 13 Sep · 19 headlines
  7. OpenAI welcomes the "AGI era" with GPT-6 Astra

    31 Aug – 6 Sep · 21 headlines
  8. An Anthropic researcher just gave us a peek at self-improving AI

    24 Aug – 30 Aug · 2 headlines
  9. Exclusive: Mirendil inks $100M+ Google Cloud deal to scale self-improving AI

    3 Aug – 9 Aug · 4 headlines
  10. Sam Altman Declares AI Singularity Has Arrived

    27 Jul – 2 Aug · 8 headlines
  11. Amazon cuts jobs in AGI unit as its AI strategy shifts again

    20 Jul – 26 Jul · 22 headlines
  12. DeepMind's Demis Hassabis predicts AGI will arrive within a few years · Digg

    13 Jul – 19 Jul · 3 headlines
  13. Is recursive self‑improvement the dawning of AI superintelligence?

    6 Jul – 12 Jul · 3 headlines
  14. Sigmatic Sciences Powers AI Scientists with NVIDIA's Latest BioNeMo Agentic Capabilities

    22 Jun – 28 Jun · 5 headlines
  15. Anthropic Declares That The Next Big Step For Humans And AI Is AI That Builds Itself Via Recursive Self-Improvement

    1 Jun – 7 Jun · 18 headlines
  16. Teamily AI Officially Enters the Singapore Market, Launching its Personal AI Agent OS -- Redefining Personal AGI Through Self-Improving AI+Human Social Network

    25 May – 31 May · 6 headlines
  17. Notable researchers join $4 billion effort to build self-improving AI

    18 May – 24 May · 6 headlines
  18. Siemens Hardware-assisted Verification Validates Arm AGI CPU for Scalable Agentic AI - Software - Siemens AG

    4 May – 10 May · 3 headlines
  19. AGI backs GRA's Publican AI platform   - The Business & Financial Times

    20 Apr – 26 Apr · 3 headlines
  20. Meta unveils first AI model from superintelligence team

    6 Apr – 12 Apr · 11 headlines
  21. Arm Announces AGI CPU For AI Data Centers

    23 Mar – 29 Mar · 17 headlines
  22. The Most Important Idea In AI: Recursive Self Improvement (RSI).

    16 Mar – 22 Mar · 3 headlines
  23. SquadStack.ai, NVIDIA Partner for Enterprise Voice AI Superintelligence

    23 Feb – 1 Mar · 9 headlines
  24. India well-placed to scale personal Superintelligence: Meta's Alexandr Wang

    16 Feb – 22 Feb · 11 headlines
  25. FinancialContent - Meta Platforms Rockets 7% as Zuckerberg Bets $135 Billion on the Future of Superintelligence

    2 Feb – 8 Feb · 4 headlines
  26. Nine Predictions for AI in 2026, from the Father of AGI – IT Business Net

    19 Jan – 25 Jan · 3 headlines
202528 chapters · 471 headlines
  1. China's AGI Pioneer and Leader Z.ai Listed on Hong Kong Stock Exchange

    5 Jan – 11 Jan · 2 headlines
  2. Amazon reshuffles AI leadership as Rohit Prasad exits, Peter DeSantis to lead AGI group

    15 Dec – 21 Dec · 8 headlines
  3. Fortanix Joins HPE's Unleash AI Partner Program to Accelerate AI Outcomes with Hyper-Secure, Turnkey Agentic AI Platform Built On NVIDIA Confidential Computing

    1 Dec – 7 Dec · 3 headlines
  4. Luma AI secures $900 million to speed up multimodal AGI push

    17 Nov – 23 Nov · 2 headlines
  5. Microsoft AI forms 'Superintelligence Team'

    3 Nov – 9 Nov · 10 headlines
  6. 800+ Scientists & Celebrities Call for Superintelligent AI Ban

    27 Oct – 2 Nov · 10 headlines
  7. Prince Harry, Meghan join call for ban on development of AI 'superintelligence'

    20 Oct – 26 Oct · 57 headlines
  8. China's Zhipu AI says full artificial superintelligence unlikely by 2030

    29 Sep – 5 Oct · 3 headlines
  9. Sam Altman and David Deutsch Agree to a New Turing Test for AGI

    22 Sep – 28 Sep · 3 headlines
  10. Meta unveils $799 smart glasses, bets on AI 'superintelligence'

    15 Sep – 21 Sep · 3 headlines
  11. Mark Zuckerberg Bets Billions On 'Startup Mode' As Meta's Superintelligence Unit Lures Alexandr Wang And Nat Friedman From Top AI Firms

    1 Sep – 7 Sep · 4 headlines
  12. In Silicon Valley's latest vibe shift, leading AI bosses are no longer so eager to talk about AGI

    25 Aug – 31 Aug · 4 headlines
  13. Amazon AGI Leader Reveals AI Talent Shortage and Junior Fast-Track

    18 Aug – 24 Aug · 6 headlines
  14. Sam Altman Explains Why GPT-5 Falls Short of True AGI Claims

    11 Aug – 17 Aug · 3 headlines
  15. Here's Why Sam Altman Says OpenAI's GPT-5 Falls Short of AGI

    4 Aug – 10 Aug · 10 headlines
  16. Meta's big AI bet: Superintelligence push unlikely to boost short-term profits

    28 Jul – 3 Aug · 9 headlines
  17. Meta's Superintelligence Team Currently Comprises Of 44 Employees, With 50 Percent Of Them Hailing From China, And 40 Percent Of The Total Headcount Previously Employed At OpenAI

    14 Jul – 20 Jul · 4 headlines
  18. apple ai: Apple AI head Ruoming Pang joins Meta as Zuckerberg boosts superintelligence team

    7 Jul – 13 Jul · 8 headlines
  19. Ilya Sutskever is CEO of Safe Superintelligence after Meta hired Gross

    30 Jun – 6 Jul · 17 headlines
  20. Meta is building an AI Superintelligence team after Llama 4's lukewarm performance

    9 Jun – 15 Jun · 3 headlines
  21. Artificial General Intelligence Market Opportunities

    19 May – 25 May · 3 headlines
  22. China's AI Strategy: Adoption Over AGI | American Enterprise Institute

    5 May – 11 May · 2 headlines
  23. Nvidia, Alphabet To Back OpenAI Co-Founder Ilya Sutskever's New AI Startup Safe Superintelligence: Report

    7 Apr – 13 Apr · 7 headlines
  24. A new, challenging AGI test stumps most AI models

    24 Mar – 30 Mar · 4 headlines
  25. EXL launches EXLerate.AI platform to drive accelerated AI business benefits at scale for enterprises

    24 Feb – 2 Mar · 3 headlines
  26. OpenAI Co-Founder Sutskever's 'Safe Superintelligence' Reaches $30 Billion AI Startup Valuation

    17 Feb – 23 Feb · 3 headlines
  27. Report: AI Startup Safe Superintelligence Aims to Quadruple Valuation to $20 Billion

    3 Feb – 9 Feb · 8 headlines
  28. Will OpenAI's Stargate Project lead to artificial superintelligence?

    20 Jan – 26 Jan · 5 headlines
202416 chapters · 238 headlines
  1. AI researcher François Chollet is co-founding a nonprofit to build benchmarks for AGI

    6 Jan – 12 Jan · 3 headlines
  2. OpenAI's o3 isn't AGI yet but it just did something no other AI has done

    23 Dec – 29 Dec · 6 headlines
  3. OpenAI launches o3 model with advanced AI reasoning and AGI potential

    16 Dec – 22 Dec · 7 headlines
  4. Most People Won't Notice When Artificial General Intelligence Arrives

    2 Dec – 8 Dec · 4 headlines
  5. ChatGPT-5 won't be coming in 2025, but superintelligence achievable today, says Sam Altman – Firstpost

    4 Nov – 10 Nov · 2 headlines
  6. OpenAI CEO Sam Altman says superintelligence could arrive in "a few thousand days"

    23 Sep – 29 Sep · 12 headlines
  7. OpenAI Co-Founder Ilya Sutskever's New AI Safety Startup Safe Superintelligence Inc. (SSI) Raises $1 Billion

    2 Sep – 8 Sep · 16 headlines
  8. Self-improving AI method increases 3D-printing efficiency

    19 Aug – 25 Aug · 4 headlines
  9. Amazon Hires Top Executives From AI Startup Adept for AGI Team

    24 Jun – 30 Jun · 4 headlines
  10. OpenAI co-founder Ilya Sutskever launches new AI company Safe Superintelligence

    17 Jun – 23 Jun · 43 headlines
  11. OneStream Unveils a Portfolio of Purpose-Built AI Solutions; Announces Enterprise Finance AI Approach to Speed AI Adoption for the Office of the CFO

    20 May – 26 May · 3 headlines
  12. Artificial General Intelligence (AGI) Market size is worth USD 27.47 Billion by 2030 with 37.5 % As Revealed In New Report

    15 Apr – 21 Apr · 2 headlines
  13. China accelerates AI development to build AI innovation center

    1 Apr – 7 Apr · 6 headlines
  14. Nvidia's Jensen Huang says AI hallucinations are solvable, artificial general intelligence is 5 years away

    18 Mar – 24 Mar · 6 headlines
  15. Nvidia CEO: Artificial General Intelligence Could Be 5 Years Away

    4 Mar – 10 Mar · 2 headlines
  16. Lenovo and Anaconda Announce Agreement to Accelerate AI Development and Deployment

    12 Feb – 18 Feb · 2 headlines
20239 chapters · 105 headlines
  1. Meta plans to buy 350K Nvidia GPUs to build artificial general intelligence

    15 Jan – 21 Jan · 5 headlines
  2. AI Scientist, Founder Of China's SenseTime Suddenly Dies

    11 Dec – 17 Dec · 6 headlines
  3. Nvidia CEO Predicts AI Breakthrough Known as AGI in Five Years

    27 Nov – 3 Dec · 4 headlines
  4. Tong Test: Innovative Method to Assess Artificial General Intelligence

    18 Sep – 24 Sep · 2 headlines
  5. Elon Musk says he told China's leaders that if a 'digital superintelligence' is created, it could take charge of the country

    10 Jul – 16 Jul · 6 headlines
  6. OpenAI forms a new team to control 'superintelligent' AIs

    3 Jul – 9 Jul · 11 headlines
  7. OpenAI leaders behind ChatGPT call for regulating AI and 'superintelligence'

    22 May – 28 May · 10 headlines
  8. The era of AI: How the Microsoft Cloud is accelerating AI transformation across industries

    24 Apr – 30 Apr · 3 headlines
  9. Keeping Singapore firms and workers at the forefront of AI advancements: Temus and AI Singapore sign MOU to accelerate AI innovation and adoption

    10 Apr – 16 Apr · 5 headlines
This weekStopping dangerous superintelligence led: 5 headlines"We won't allow SI to be stifled": Trump opposes global regulation of superintelligence at 33rd APEC summit Gone quietNothing has gone quietEvery question that was loud is still being written about.

Part of Argus Overwatch's investigations: a topic followed over months, as the record wrote it. Stories link to their pages and the articles behind them; names link to their own pages; claims link to the fact-checkers who rated them. Nothing on these pages is projected; a quiet week is drawn as a quiet week.