At the beginning of 2026, Elon Musk posted five words on his social network: “We have entered the Singularity.” Months later, on a podcast, Sam Altman remarked that we were “like, inside the singularity.” Two of the most powerful figures in the technology industry were announcing, with the casualness of someone discussing the weather, that humanity had crossed one of the most debated thresholds projected for the future. And for me, not being a technologist, what was striking was not whether they were correct. It was that no one—neither they, nor those who applauded them, nor those who criticised them—could state how one would verify that such an event had occurred.
One must be fair to both, as they did not say the same thing, nor with the same degree of certainty. Musk was categorical and extremely brief: “we have entered,” full stop, as one stating a fact. Altman was more cautious, and the “as if” he included, “as if within the singularity,” is not an oversight, but a nuance: in his own texts, the singularity is not an instant one crosses, but a gradual process which, according to him, has already begun but has not yet culminated. He speaks of being close to constructing a superintelligence, not of having already constructed it. Thus, even between the two announcers, there is no agreement on what is being announced. One proclaims a fait accompli; the other, an ongoing ascent. And that discrepancy, among those who should know best, already speaks volumes about the problem.
The problem is that “singularity” is not a measurable magnitude, but a word that has spent decades naming different phenomena. The mathematician John von Neumann, as recalled by a colleague in the 1950s, spoke of a point of technological acceleration beyond which human life as we know it could not continue. In the 1960s, another mathematician imagined it as an intelligence explosion: a machine sufficiently capable of designing even more capable machines, and so on, without restraint. In the 1990s, a writer defined it as the emergence of intelligences superior to that of humans. Later, it was popularised as a date, around 2045, when the curves of progress would soar. These are related ideas, yet they are not identical, and each would place the threshold in a different location. To ask if “we have arrived” without specifying which one we are referring to is akin to asking if we have travelled far without stating the destination.
When someone says “we have entered the singularity,” they perform an act that appears to be a verification but is not. For it to be one, elements that are currently absent would be required: a definition of exactly what it is we claim to have reached, indicators that it has been reached, a threshold marking the before and after, and, above all, a condition that would tell us what we would need to observe to conclude that we have not yet reached it. Without that final piece, the assertion carries no risk: whatever the state of the world, one could always claim it fits. And an assertion that fits everything has ceased to assert anything verifiable. I do not claim it is false, mind you; that would be a different error. I claim it is not yet a demonstration, but a declaration: it holds value through the authority of the one who makes it, not through evidence that others may review.
This difference, between declaring and demonstrating, I know well from another field, my own, and that is why the case feels familiar. In art, the exact same thing occurs. For decades we have debated whether something is art, and a convenient answer has been: it is art if the art world says so, if it is consecrated by a museum, a critic, a fair. But that answer suffers from the same vice as Musk’s statement. For an authority to declare “this is art” is not a criterion; it is a declaration. It holds value through the position of the one who issues it, not through a property that anyone can verify. I have spent years attempting to formulate a criterion to distinguish when an artistic event occurs and when it does not, and I know from experience how difficult it is, how much it costs for a criterion to be both precise and honest regarding its limits. Precisely for that reason, I recognise from afar when there is none. And in “we have entered the singularity,” there is none.
It is not a matter of denying that technology is changing rapidly, nor of dismissing anyone as naive. We may indeed be facing a profound transformation; I have no way of knowing, and that is precisely the point. No one does, because we have not agreed upon what would constitute such knowledge. What these statements offer us is not data, but a diagnosis endorsed by the status of the speaker. One must distinguish between the two, for the confusion between them is longstanding and extends beyond technology: it is the mechanism by which an authority—be it in science, art, or politics—replaces the evidence they lack with the weight of their own reputation. Before debating whether we have crossed the threshold, we must first agree on where it lies. Until we do, anyone claiming we have already crossed it is not informing us of a fact. They are asking us to believe them.
On the open conversation
This text takes two recent and verified statements—Elon Musk’s assertion that “we have entered the Singularity” and Sam Altman’s claim that “we are, like, inside the singularity”—as a starting point to examine the distinction between declaring that an event has occurred and possessing a criterion to demonstrate it. I respect the asymmetry between the two (Musk was categorical; Altman, more gradualist) and I do not argue that they are mistaken or dishonest: I argue that “singularity” encompasses disparate definitions and that, without agreeing on which one applies, the assertion is not verifiable. I connect this to a vice I have encountered in art theory: confusing an authority’s declaration with proof of the matter. If anyone wishes to intervene from the perspective of the philosophy of science, epistemology, or art theory, this notebook remains open.
Sources
Elon Musk: “We have entered the Singularity” (X, 4 January 2026) and “We are in the beginning of the Singularity” (1 February 2026).
Sam Altman: “We’re now, like, in the singularity” (Relentless podcast, 25 July 2026); “The Gentle Singularity” (blog, June 2025: “We are past the event horizon; the takeoff has started”, and simultaneously “close to building digital superintelligence”); “The Merge” (2017), where he defends the singularity as a gradual process.
Adam Shepherd, “Elon Musk and Sam Altman claim we’ve reached the AI singularity. But how would we even know that happened?”, Live Science, 17 August 2026.
Genealogy of the concept: S. Ulam recalling J. von Neumann (Bulletin of the AMS, 1958); I. J. Good (1965); V. Vinge, The Coming Technological Singularity (1993); R. Kurzweil, The Singularity Is Near (2005); D. Chalmers, “The Singularity: A Philosophical Analysis” (2010).
On falsifiability and testing: K. Popper (with subsequent nuances from Duhem, Quine, Kuhn, and Lakatos).