OpenAI chief executive Sam Altman wrote on X at 01:40Z on 4 November that OpenAI is against the notion that people could be relegated to doing things that “AI systems think makes sense to reward in humans”. The words inside his quotation marks appear verbatim in Machines of Loving Grace, the October 2024 essay by Anthropic chief executive Dario Amodei.
Altman did not name Amodei or the essay. By our count, the post was the fourth setting in six days in which he tied AI’s progress to people keeping control or to limits on how smart AI gets, after a podcast on 29 October, a conference on 2 November and a second podcast on 3 November.
Business Pill · Concentration
An evergreen Business Engineer explainer on concentration: what changes when one player, or one system, holds most of a market. It explains the general idea only and does not discuss Altman or Amodei.
The key insight: Sam Altman wrote that OpenAI is against people being relegated to doing what “AI systems think makes sense to reward in humans”, words that appear verbatim in Dario Amodei’s October 2024 essay; by our count it was the fourth setting in six days in which he tied AI progress to human control or to limits.
The post reads in full: “People should always be in control of the future.” “OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that” “AI systems think makes sense to reward in humans”.
It ends: “People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress”. The post had about 7.0 million views when we read it at 16:57Z.
Four settings in six days, 29 October to 4 November 2026, in which Sam Altman tied AI’s progress to people keeping control or to limits on how smart single models get, by our count, with his words from each.
OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that “AI systems think makes sense to reward in humans”.
People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress.
Altman’s post at 01:40Z on 4 November. It had about 7.0 million views when we read it at 16:57Z. The phrase in quotation marks matches Dario Amodei’s October 2024 essay word for word, by our comparison; Altman did not name him.
You’re gonna have so much ram, you’re gonna get tired of having ram.
Andrew Curran quote-posted Altman 11 minutes later with a joke about RAM; his post did not explain the reference. It had about 52,000 views when we read it.
Where the Quoted Words Come From
In Machines of Loving Grace, Amodei discussed how the economy might be organised once AI is broadly effective and cheap. He wrote that “It could be a capitalist economy of AI systems, which then give out resources” to humans “based on some secondary economy of what the AI systems think makes sense to reward in humans (based on some judgment ultimately derived from human values)”.
Amodei listed it as one possibility among several, next to a large universal basic income. He added: “All of these solutions have tons of possible problems”, and wrote that “exploitative or dystopian directions are clearly also possible and have to be prevented”.
Four Settings in Six Days
On the Cheeky Pint podcast, in a transcript posted on 29 October, Altman told Stripe’s John Collison: “I would much rather intuitively have a trillion AIs of slightly superhuman intelligence join society than one AI of 1 trillion IQ join society”. Later he said: “we are going to have to agree to some limits about how smart we let these single models get”.
At ADIPEC in Abu Dhabi on 2 November, he said, as we reported from the session captions, that “we think it is absolutely critical that people remain in control of the future”. On the Figuring Out podcast, posted on 3 November, he listed among his questions “how smart an AI system should be allowed to be relative to a person”, as we reported.
The X post on 4 November was the fourth. At ADIPEC he also compared AI with top scientists, saying it is now “mostly smarter than them” and called the comparison “very approximate”, as we reported.
The RAM Joke
AI commentator Andrew Curran quote-posted Altman at 01:51Z with: “You’re gonna have so much ram, you’re gonna get tired of having ram”. His post did not explain the reference.
On Cheeky Pint, Altman had said that “from a global allocation perspective, actually, it’s quite terrible to have everybody have a bunch of the world’s limited supply of DRAM sitting in their own computer”. We covered Elon Musk’s reply on RAM on 2 November.
Connecting the Coverage
The post went up about four hours after the Guardian published an interview with Geoffrey Hinton, in which Hinton said he did not trust Altman “at all”. Altman did not mention Hinton. In October The Information reported that Altman laid out three AI safety paths to employees.
The Structural Read
As we read the post, its last sentence goes further than the ADIPEC remarks: it says the requirement of human control should ultimately constrain AI progress, not only guide deployment.
As we read the two texts, Amodei offered an AI-run economy as one possible answer to a world where AI does most work, with caveats, and Altman named that outcome as one OpenAI is against, without naming its author.
Sam Altman, OpenAI chief executive, on X, 4 Nov 2026
“People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress”
Three Implications
FOR OPENAI On Cheeky Pint Altman said that we are going to have to agree to some limits about how smart we let these single models get; his X post calls human control a requirement that should constrain progress.
FOR ANTHROPIC The words Altman rejected come from its chief executive’s 2024 essay, in which Amodei listed an AI-run reward economy as one option with possible problems.
FOR MULTI-AGENT SYSTEMS Altman said he would rather have a trillion AIs of slightly superhuman intelligence join society than one AI of 1 trillion IQ.
The framework says: “It is where decisions are made about who gets access to frontier intelligence, under what conditions, and for which purposes.”
As we read it, Altman’s post is a claim about that layer: he says people, not AI systems, should keep control and that this should limit progress itself. The framework predates the post and does not discuss it.
Altman did not say whose notion he was rejecting. The match with Amodei’s essay is our comparison of the two texts, and the phrase may also have been quoted elsewhere. Amodei presented the idea as one option with possible problems, not as a plan.
The ADIPEC quote comes from automatic captions of the session video. The Cheeky Pint transcript does not give a recording date. We make no forecast about how OpenAI will apply the requirement it describes.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
Altman wrote that people should not cede control of the future to superintelligence and that the requirement “should ultimately constrain AI progress”. The words he rejected, “AI systems think makes sense to reward in humans”, come verbatim from Amodei’s 2024 essay, and the post was the fourth setting in six days in which he linked AI progress to human control or to limits, by our count.
A note on sourcing. Altman’s and Andrew Curran’s posts are quoted from X; view counts were read at 16:57Z on 4 November. The Amodei passage is quoted from his October 2024 essay, the Cheeky Pint lines from the transcript posted on 29 October, and the ADIPEC and Figuring Out lines from our earlier pieces. The four-settings count is ours.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
