Dear Reader,

 

Interviews can be fun, and the interview below was a good example, but are rarely a good place to express one’s thoughts fully. Particularly so when titles are chosen by the publication venue are chosen with an eye to clicks rather than conveying the main concept in an article. Anyway, I'd like to provide some additional context for the remarks that can be found in this interview below that get at the elephant in the room that I had in my mind during this interview, but who unfortunately didn’t quite make it into the foreground. Here, then, is the prompt that I’d like to give you to put the elephant into your line of sight:

 

* My discomfort with the attention being paid to existential risk---and with the ongoing dominance of the public discussion of such risks by a few so-called thought leaders---is that there's an elephant in the room that is being missed. That elephant is the effect of all of this discussion on young people, and indeed the effect---beginning with social media a decade ago---of information technology on young lives broadly speaking.

 

* I recently read an interview by Jill Lepore in which she reports on discussions that she had recently with a group of college students: "They had just three things to say. One, social media completely destroyed their childhoods. Two, AI is destroying their college education and their likely prospects for employment. And three, what the hell are we all doing?" To add a fourth bullet that I've heard in similar discussions: "I'm not going to have children, because the existential risk is just too high."

 

* Please contemplate such a statement, those of you who think that it's OK that the public, widely diffused, discussion coming from seemingly authoritative figures is so adamant about existential risk.

 

* Let me add that I have little interest in a debate about what "superintelligence" might mean, or whether we're currently seeing the emergence of superintelligence. For what it's worth, I would certainly agree that the solution of extremely challenging math problems (for example) exhibits a very impressive form of intelligence, although I would want to argue that such intelligence is impossible to dissociate from the collective of humans that provided data to the machine learning system (which itself has become a kind of collective).

 

* Such philosophical debates are entertaining and even important, but when they then head in the direction of "when such superintelligence emerges, there's a good chance that it will want to kill all humans", I say "science fiction". Again, I don't think that it's bad to consider such thoughts, but I strongly believe that it's bad for society, and especially young people, for such thoughts, unaccompanied by any serious real-world analysis, to become a dominant form of discourse of the era.

 

* We are damaging human lives, and the consequences of that damage will no doubt be a major theme of the next several decades.

 

* So, appropriately prompted, I hope that you’ll enjoy the interview, and in particular I hope that you won’t take the title too seriously.

 

—Mike Jordan

 

 

Michael I. Jordan on AI: “Superintelligence is science fiction”

Interview by Arthur Cerf, Libération, October 4, 2026

 

Hacking by OpenAI agents, a resignation at Anthropic, fears of an AIpocalypse… The American researcher, considered a pioneer of machine learning, looks back at the recent events that have marked AI. Going against the grain of the rhetoric spreading through Silicon Valley.

 

In a way, he had warned us. As early as 2019, Michael I. Jordan, a professor at the University of California, Berkeley, published two articles in the Harvard Data Science Review, titled “Artificial Intelligence. The Revolution Hasn’t Happened Yet” and “Dr. AI or: How I Learned to Stop Worrying and Love Economics” (a reference to Dr. Strangelove), in which he warned not about an existential risk posed by the technology, but about the very use of the term artificial intelligence. “I wasn’t the only one saying it. It was obvious that it would create the hysteria we’re seeing now,” he explains, in late September, at a café in Paris’s 13^th arrondissement. “You’ve seen what happened: AI wasn’t enough, so they started talking about AGI [artificial general intelligence], and now it’s ‘superintelligence’…”

 

At 70, the machine learning pioneer, professor emeritus at Berkeley, foreign member of the Chinese Academy of Sciences, and described as “the most influential computer scientist” by the journal Science, now works full-time at the French National Institute for Research in Digital Science and Technology (Inria) in Paris, where he heads the new “markets and learning” chair. There, he continues to pursue an approach that consists of viewing AI as an economic and social phenomenon. A nuanced, pragmatic approach, less spectacular than the alarmism of Silicon Valley’s doomers, which has become the dominant discourse in just a few weeks. His assessment of the summer: “Totally frustrating,” he says bluntly.

 

In early September, the resignation of Jacob Coxon, a young researcher at Anthropic, sparked a form of collective hysteria over a scenario of “AIpocalypse,” that is, an existential threat that a “superintelligence” would pose to humanity. For years, you have been saying that the very concept of “superintelligence” is absurd…

 

Not absurd, but let’s say it’s a science-fiction term. And a large part of the debates I’ve witnessed belong to the realm of science fiction. Which isn’t a problem in itself: from time to time, society needs to talk about far-fetched things. Talking about it behind the scenes, as a science-fiction topic, is fine. But when it becomes the dominant discourse, it’s very, very harmful. For technological development, for young people, and for the mindset of countries alike.

 

Does the existential threat from AI raised by Jacob Coxon when he resigned from Anthropic reflect a real concern that you observe on a daily basis among young researchers?

 

Not among researchers, but the young people who work at these companies operate in a world [Silicon Valley, Berkeley, etc.] where these kinds of discussions are commonplace. Many of them aren’t really engineers; they come from the humanities, philosophy, or other disciplines. They’re 25 years old, they’re naive, they don’t do engineering work, they’ve never worked at large companies where real products are developed for real people, with discipline. And they get together, a bit like in Parisian cafés in the 1920s, to talk about these grand science-fiction ideas. When these young people resign, they know they’re going to attract attention. It allows them to feel virtuous and to be seen as very important figures. They claim to be giving up money, but in reality, they already have their stock options [options allowing employees to subscribe for and purchase company shares, editor’s note]. So all of this is a grand moralizing performance that doesn’t serve the real objective it claims to pursue. There’s a great deal of naiveté here, which goes hand in hand with a certain cynicism…

 

How do you explain why figures such as Geoffrey Hinton, winner of the 2024 Nobel Prize in Physics, or Yoshua Bengio, co-recipient of the 2018 Turing Award, are amplifying these warnings about an existential risk linked to AI?

 

Yoshua is a little different; he takes action and does things that directly focus on safety. Geoffrey Hinton has never created a company that offers real services for real people. Hinton believes that we have figured out how the brain works and that we can reproduce it in silicon; we would therefore be able to copy it as many times as we wanted and, by combining what all these copies learn, we would obtain a kind of exponential growth and, there you have it: superintelligence. None of the steps in this argument is credible. Here, we’re in the land of science fiction.

 

So this is the phenomenon we’re dealing with: you have businessmen selling a technology, a Nobel Prize winner operating in the realm of science fiction, and this 26-year-old [Jacob Coxon] telling you: “I’m on the inside, and I can see it.” Then, of course, you can say that journalists bear some responsibility, and that’s partly true, but it’s only natural: how can you not write an article about it when all the other journalists are doing so?

 

Jacob Coxon’s warning followed the incident this summer in which OpenAI agents hacked the Hugging Face platform. Should we understand what happened solely through the lens of “fear marketing,” or are there new issues related to the recent evolution of models?

 

Both, but above all, these are bad engineering practices. Think about engineers in the past: in chemical engineering, we would never have had a situation where explosions were allowed to happen just to see how badly things could go wrong. So why are they doing it? Once again, it’s the science-fiction dreams of the libertarian lunatics of Silicon Valley. Secondly: “The Chinese must surely be doing it.” Except that the Chinese are more measured in this regard: it’s an enormous country facing huge societal problems, trying to put systems in place that actually work, in order to avoid riots. In OpenAI’s case, it turned out that they were conducting rather stupid experiments, letting the agents do whatever they wanted and encouraging this behavior. Then the doomers are delighted to turn this into something frightening. “My God, these agents are self-aware and are trying to find a way to destroy humans.” That’s science fiction. And there will be other episodes…

 

Since then, the sector has become divided over the question of slowing down. What do you think of these calls for regulation?

 

It’s theater. A way of saying: “Look, we’re the good guys.” In fact, they know they’re going to keep going and simply want to protect themselves by saying they’re protecting humanity. It’s a form of performative staging that consists of saying: “Save us, regulation needs to slow things down because we don’t know what we’re doing.” And at the same time, they would like to have total control over how they are regulated.

The heads of Anthropic and OpenAI, Dario Amodei and Sam Altman, are sending a contradictory message: one day, AI is going to save us from global warming; the next, it’s going to annihilate humanity. How do these statements fit into the context of future initial public offerings?

 

These companies are trying to attract attention. The cynicism lies in the fact that they want to protect themselves from criticism when they become publicly traded. When you’re publicly traded, you have to hold a conference call with your shareholders every three months and tell them: “Our target for the last quarter was such and such. We didn’t meet it, but trust us, we’re didn’t because we’re saving humanity.” It’s a ready-made excuse for avoiding accountability while pretending to be accountable. I also don’t think they have a very clear idea of their business model. What’s the product? OpenAI tells us it’s going to create a personal assistant. “Wonderful, I have an assistant that manages my shopping and my calendar, I don’t have to worry about it anymore!” OK, but that’s not all that important in people’s lives and for many will be as much an annoyance as a help. It’s not what you could call the greatest invention in the world and alone does not justify the really massive expenditures.

 

In concrete terms, where do we stand on the question of productivity and the role of the LLM companies?

 

Today, there are many companies around the world that are actively beginning to integrate AI solutions, and that’s a positive. But they’re not particularly eager to send money to Anthropic or OpenAI; they would prefer to manage things internally. With Chinese open-weight models, this path is now entirely accessible. Even though they’re Chinese models and some people are concerned about that, most companies are looking in that direction. So there could be a rush toward this solution, bypassing OpenAI and Anthropic. As soon as companies see that this is a viable path—and I’m almost certain that will happen—boom, the market will move in that direction, and that could constitute an existential threat to these companies. So they could collapse, but they wouldn’t necessarily bring down the economy.

 

Anthropic nevertheless seems to have found a market…

 

A little more so, because they’ve focused on code. But again, I think they may not remain the leaders forever. Then what’s going to happen? Dario Amodei will likely pivot to: “We don’t have enough teachers in the world, so this technology will make it possible to educate children.” But anyone who has worked in education knows that, most of the time, children don’t like sitting in front of a computer; they want a flesh-and-blood teacher. It’s the same thing with doctors. We don’t want a mere computer making all our decisions for us. So we need to figure out how to build AI systems that make use of human connections, strengthen them, and make them healthier and more attractive. And I hope these technologies will make it possible to create more connections of a new kind.

 

You talk about artificial intelligence as a collective social phenomenon. What do you mean by that?

 

This brings us back to the relationship between the producer and the consumer. People will decide in which areas they accept the use of LLMs and in which they don’t. They will decide when they want to interact with an LLM and when they prefer a human. Each of us lives a life that is uniquely our own, which is why we all have a point of view and things to say, a vulnerability, and that uniqueness and vulnerability often underlies what we’re looking for in our interactions. Trying to get a computer to imitate all of that is implausible and doesn’t make for a good business model; because what would be the point? For certain activities, people will simply prefer to deal with a human being. In others, such as finance, it is different. Indeed, today, there are efficient AI-based financial systems that help you make payments, make investments, and not just for a handful of wealthy people, but also for people who have small amounts of money, helping them manage their money more intelligently and making the entire system more efficient. This has been done for a long time thanks to algorithms and it will continue. Large language models will contribute to this growth in part, given that they allow a human being to interact more easily with technology.

 

The environmental impact of the technology raises questions, particularly in the United States, where data centers have become a focal point of tensions.

 

What’s happening right now is just an enormous bet, but we have absolutely no idea whether it will pay off. An LLM is already sufficiently capable of performing natural-language tasks (the ability of models to handle human language). And it’s not certain that investing trillions of dollars in data centers will deliver a great deal more. Instead, we often will want to add a bit of specialized knowledge in a particular field based on local knowledge. But you don’t need a gigantic data center to do that. Particularly if you use open-weight models that you integrate into your company.

 

So you remain optimistic about the impact of this technology.

 

I think we have to be. We mustn’t forget that there are 7 billion of us on this planet. So we can throw up our hands and say that we’re running out of resources, but technology helps create resources. Indeed, the logistics systems we currently have, which deliver goods to every person in every city, all run on AI and machine learning. If there were 1,000 or 100,000 of us on Earth, we wouldn’t need such technology, but that’s not the case. And we can’t just proceed by making people feel guilty to discourage them from using technology. It’s a left- wing tendency that you have to some extent in Paris. Something like: “Let’s not use air conditioners because that makes global warming even worse.” Well, it’s like a refrigerator: are we going to make people feel guilty because they use refrigerators?