Alexander Wissner-Gross's new theory of physics is built on one sentence, and he says it produces one number: about 1/137.036. That number is the fine-structure constant, usually written with the Greek letter alpha. It measures how strongly light and electric charge interact. Physicists have measured it with great precision, but the standard theory of particle physics cannot explain why it has that value. It simply takes the measured number as an input.
Wissner-Gross, a computer scientist, founder of Reified and co-host of the Moonshots podcast, introduced Chain Theory on Saturday, October 24, 2026. He calls it a research program to rebuild fundamental physics from quantum information. On the Moonshots episode recorded October 27, host Peter Diamandis, founder of XPRIZE, called him the show's "very own in-house ASI" (artificial superintelligence) and handed him the science segment. Wissner-Gross argued that the constant "falls out with no extra inputs directly from this axiom." He also said the result could be proved wrong within months or years, and that the AI tools he used needed a lot of steering.
One sentence about entangled pairs
Diamandis opened with a clip from an explainer video that set out the clues behind the theory. The first, as the clip told it, came from Jacob Bekenstein's 1981 work on black holes, which put an upper limit on how much information a bounded region of space can hold. Physicists usually treat that limit as a fact about gravity. In the clip's account, Wissner-Gross notes that it also covers the matter inside the region, while today's theories describe matter with fields that have infinitely many possible settings at every point. His answer is a world built from a finite number of tiny quantum systems. A qubit has two settings, a qutrit three, and the general case is called a qudit.
The second clue was a pattern. Elementary particles behave like points with no inner parts, yet they come in three "generations" that carry the same set of charges. The clip argued that a repeating pattern like that has historically pointed to smaller parts, so the parts may exist but be locked together too tightly to pull apart.
The theory pairs these systems up and entangles each pair "maximally." Entanglement links quantum systems so their measurement results are correlated in ways no ordinary connection allows, and a maximally entangled pair is as tightly correlated as quantum mechanics permits. The clip gave the reason for the "maximally": a weaker link would need a dial set to a number with infinitely many digits, and a finite world has no room for one. The resulting axiom is 18 words long: "Nothing exists but finitely many, maximally entangled pairs of finite dimensional systems, some shared, known only by distilling them."
In the clip's explanation, "distilling" means joining two systems into a perfectly entangled pair in one step, the same move quantum networks use to pass entanglement along. Take two pairs and make one joint measurement on the two inner ends, and the two far ends become a pair. In a laboratory that step works only some of the time. In Chain Theory it is the only thing that ever happens, and nobody observes the network from outside: learning about a link means linking to it. "So you, me, every measuring device, and every AI that will ever read this, we're all chain mail," the clip's narrator said.
Wissner-Gross said the name contrasts his theory with string theory, a popular candidate theory of quantum gravity, which assumes continuous degrees of freedom. Chain Theory's are finite and discrete. The name also fits the picture of a physical chain made of links. Under the theory, he said, electrons, quarks and photons are "basically just small networks" of these finite quantum systems.
A number with nothing tuned by hand
The clip compared the starting points. The Standard Model of particle physics needs 17 particles and about two dozen measured numbers before it can predict anything. Wissner-Gross put the count of free parameters, the numbers that must be measured and plugged in by hand, at 18, or 19 counting one convention, and that is before neutrino masses. Many are known to 10 or 11 significant figures, he said, "but they're total mysteries."
His first result is the fine-structure constant at very low energies, which physicists call alpha zero. The clip said his value matches measurement to 10 digits. Later in the episode, Wissner-Gross read the exact expression aloud: 79 plus 48 times the square root of 2, plus 42 times the square root of 3, plus 3 times the square root of 6, all divided by 24 times 36 squared. The expression is built only from whole numbers, and one of them is 42. "So, Douglas Adams and 42, maybe you were right," he said, nodding to the novelist who made 42 the answer to life, the universe and everything.
He described other constants as still on the way. "Masses fall out of it," he said, along with other coupling constants and some cosmological parameters. "Every preliminary indication that I've seen is that it looks plausible that every single free parameter of the standard model could fall out of this," he said, while calling these "early days."
Forces as kinds of links
Asked by Salim Ismail, founder of OpenExO, whether the goal was to describe all particles, forces and spacetime from information and entanglement, Wissner-Gross said: "That's the aspiration, yes." He laid out what he called a mapping between link types, set by the number of settings on each end of a pair. A qubit linked to a qubit (2-2) appears to map to spacetime. A qutrit linked to a qubit (3-2) maps to electromagnetism, which is where alpha comes from. In his mapping, 3-3 gives the strong nuclear force, 4-2 the weak nuclear force, 4-3 the Yukawa couplings and 4-4 the Higgs. His introduction makes the same assignments.
Dave Blundin, founder of Link Ventures, asked whether that opened another test: might a 5-4 link be an undiscovered force? "It's possible," Wissner-Gross said. He added that the theory treats the Yukawa couplings and the Higgs as equal to the other forces. Physicists normally teach them as connections between matter particles and the Higgs field, and as the Higgs itself. In Chain Theory's view, he said, "we sort of confused the fifth and sixth forces with the mass spectra."
Diamandis asked about dark matter. Wissner-Gross pointed to work by Polish quantum information researchers, the Horodeckis, who showed in 1998 that some entangled states cannot be distilled into perfect pairs, even when many copies are available. Physicists call this "bound" entanglement, and the Horodeckis' example used two three-dimensional systems. Wissner-Gross compared such a state to a prime number that cannot be broken into simpler factors. Dark matter interacts through gravity and through none of the other known forces. He said Chain Theory has a "dark interaction" that is undistillable, and that with high enough dimensions such links "would be beautiful candidates for dark matter."
What it might unlock
If the axiom is correct, Wissner-Gross said, "the bottom of the pool, as it were, is a lot closer than many would think." Textbook physics treats the world as continuous down to the Planck scale, many orders of magnitude beyond the reach of cosmic rays or particle colliders. Chain Theory, he said, suggests the physical limits of length, time and energy, "for some definition" of each, may be much closer to today's technology. He hedged heavily: "This is a very new theory. Parts are underdeveloped."
Emad Mostaque, the episode's guest, called it "the most simulation friendly physics theory out there." Wissner-Gross agreed that it looks "suspiciously" like what physics would look like if the world were built on a quantum computer.
Diamandis, calling on the "Space Cadet" in him, asked about inertialess drives and anti-gravity. "I would love that more than anyone. I don't know yet," Wissner-Gross said. He thinks any "Star Trek sci-fi tech" will require a breakthrough in fundamental physics, perhaps one like Chain Theory, and summed up his position as "cautiously optimistic." Ismail's response: "I'm going to have to take some mushrooms and then come back and listen to this segment again."
How much the AI did
Blundin asked how far along the work was before AI. Was it 80, 90 or 99% done before a model "pushed everything over the finish line"? Wissner-Gross said he had been thinking about these questions for decades. He earned a physics PhD at Harvard in 2007 and later published on causality, computer science and, in 2010, "causal and entropic descriptions of intelligence." He said he does use what he calls superintelligence, but the systems he has access to cannot reach something like Chain Theory on their own. "It requires an enormous amount of steering to make progress," he said. "Chain theory was, I'll call it my intellectual brainchild."
Where AI helped most was Lean, a programming language in which a computer checks every step of a mathematical proof. Chain Theory comes with a Lean proof that runs from the axiom to the number, Wissner-Gross said. The proof does not establish that the number belongs to the photon, the particle of light. That correspondence with real electromagnetism is still a conjecture, set out in the paper but outside the proof. Working with the models, he said, is like running "an entire department of very talented graduate students that require an enormous amount of steering and taste at the moment." He doubted the comparison would hold more than a few more months and called it "probably inevitable" that the steering and taste will be automated too.
Blundin said that statement defines the current moment in history: a scientist can be "massively more efficient and is not yet completely irrelevant." The exchange ended on the idea that this may be the period of the last human discoveries, so people had better make one.
Released before peer review
Asked by Diamandis for his view, Mostaque said he had not yet read "all 700 pages" but praised the approach. String theory, he argued, has built "a whole edifice" of elegant mathematics on a presupposition, but "it's not even wrong in that you can't prove it's wrong." He said Wissner-Gross's base in information theory is "more solid than string theory," and "at least this can be tested." He noted that measurements of the constant from Paris and Berkeley differ by a little more than a standard deviation. Mostaque described the release, made publicly rather than through peer review first, as fitting a time when everyone has AI that can check results: "who is the peers now?" He added that "we have to be cautious of getting overconfident about these things."
Wissner-Gross set his release against what he called AI companies' "slop drops," bulk solutions of existing open math problems that have drawn complaints from mathematicians. "It is absolutely essential that there be new conjectures, new insights, not just bulk solving of existing open problems," he said. He described physics as being in "a bit of a depression," with the search for ever simpler explanations going into reverse around the 1970s, and argued that clues to deeper physics are "hidden in the constants."
He has published many refereed papers, he said, but this one "had an urgency to it that required immediate publication," because refereeing can take a year and a half. Alongside the letter on alpha he released a compendium he put at about 800 pages, covering the Standard Model, general relativity, cosmology and other areas. He expects the work to appear on the arXiv preprint server soon and to be submitted to a physics journal, where he thinks quantum information researchers will probably referee it.
The test he points to is measurement. "So it's falsifiable," Wissner-Gross said. He expects experiments, "potentially in the next few months, potentially the next couple of years," to measure the next significant figure or so of alpha zero "to decide, is this actually a correct prediction or not?"