Eric Weinstein on Geometric Unity: One Canvas, Fourteen Dimensions and a Theory of Everything

Lex Fridman opens by asking how Eric Weinstein felt after posting his Oxford lecture on Geometric Unity, the product of more than thirty years of work. Weinstein replies that the great sin in academia is to appear unserious or deluded and to ignore the standard practices everyone has signed up for. He knew many colleagues would be angry that he went outside incremental publication and departmental norms.

He says that when he was inside the system, its most senior and respectable members seemed to him functionally insane, like a health agency head making recommendations they do not believe in. In his view the physics world was "really crazy", while mathematics was less so but strict and dogmatic.

Weinstein explains that for years he kept the number fourteen as a closely guarded secret: four dimensions of space and time plus ten extra dimensions of rulers and protractors. He talked to almost no one, and compares telling friends to "coming out", saying many people have closets unrelated to sexual orientation. He adds that he pretended to care about other topics as a cover for what he really cared about.

He draws an analogy with artificial intelligence, where large departments build narrow tools but few work on intelligence itself. Because nobody knows how to attack the big question directly, people work on tangentially related programmes, and he says he left the road of tangents. He recalls asking his department about a puzzling line in a book and finding nobody could interpret it correctly.

Fridman recalls watching Weinstein's livestream, in the early hours, around April 1st, and sensing a serious decision being made. Asked why he released it now, Weinstein says he does not think the date was comedy. He believes the Covid epidemic is "the end of the big nap", and notes he had tried this seven years earlier at Oxford, when it was too early. In the intervening decade he had moved to the internet, which is where his podcast, The Portal, lives.

The Oxford experience was disappointing, he says, because of two separate issues: whether the idea would actually be heard and explored, and whether he would become disconnected from the work through ridicule, immediate improvement of it, or claims that it is derivative of a 1957 paper. He likens it to a policeman selectively enforcing obscure regulations against you.

Weinstein says he does not believe a final theory can be reached within the political economy of academia, because the selective pressures on pay, freedom and mutual criticism will eliminate anyone likely to make real progress. Fridman raises Andrew Wiles, who worked in isolation on Fermat's Last Theorem. Weinstein replies that Wiles' original proof was, as he recalls, unsalvageable, and that a second proof with Richard Taylor carried the day.

Weinstein says that people wrongly assume he implicitly agrees with the community on how things should be handled, and that his critique is large in scale. He distinguishes constructive from destructive criticism and points to physics' record on string theory, as well as mask policy and the financial crisis, asking why those who failed for 35 years keep their prestige.

He also cites Dirac's argument that beauty can be a guide in physics. In Weinstein's telling, Dirac said that a mismatch with experiment can be a fine-tuning issue and not fatal, as when Heisenberg objected to Dirac's theory over proton and electron masses.

At Fridman's request for a version "for idiots", Weinstein says the goal of Geometric Unity is to start from something as close to a mathematical nothing as possible, a blank canvas. Current physics has two sectors: general relativity, where gravity is the star, and the standard model, which covers the three non-gravitational forces and matter. Photons, gluons and intermediate vector bosons feature in the latter.

He defines a theory of everything as one where questions beyond it are no longer mathematical in nature. Saying spacetime is a four-dimensional manifold only supplies a place for calculus and linear algebra to interact, like ordering a canvas. The hard part, he says, is whether the canvas can produce the artist, paints and brushes, and he mentions the image of hands that draw themselves.

In the usual picture, Weinstein says, general relativity is the theory of the canvas, optionally warped and with length and angle. The standard model, with its symmetries SU(3) x SU(2) x U(1), sixteen particles per generation in three copies, and the Higgs field, is imported on top like foreign libraries. This double origin story was the target of earlier attempts such as Kaluza-Klein theory, and Weinstein says grand unified theories tried to tidy the second origin without removing it.

He adds a speculative aside that we are the AGI: stories about simulation and about a self-aware program taking over are, he says, the same story, and learning our own source code could be our "Skynet moment". He stresses that Geometric Unity is speculation, not accepted physics.

Weinstein uses the programming idea of technical debt to explain his construction. If you cannot choose one set of rulers and protractors, take the set of all possible ones. Over a four-dimensional spacetime that adds ten variables: four rulers plus six possible angles, giving 4 + 4 + 6 = 14. The result is a fourteen-dimensional space he called U14 in the lecture and now calls Y14.

The cost is that he now has to explain why we perceive only four dimensions, which he says means sampling a four-dimensional slice, a "section of a bundle". The payoff, he says, is that spinors, the fermions, arise naturally on a space that nearly has length and angle, without extra internal quantum numbers. Because Y14 is a space of measurements, it almost carries a measurement structure itself.

He describes spinors as the first wildly surprising object hidden in a manifold, "like buying a house and finding a panic room". They need 720 degrees of rotation to return to normal, which he illustrates with the Dirac belt trick and a glass of water, the "Philippine wineglass dance". Dirac found them while taking a square root of the Klein-Gordon equation, and Weinstein calls them among the most profound aspects of human existence.

Asked what the manifold is, he says it is proto-spacetime: a four-dimensional manifold before Einstein adds the pseudo-Riemannian metric, which is equivalent to length and angle, a symmetric four-by-four matrix.

Fridman, imagining a book called Geometric Unity for Idiots, asks whether gauge theory and spinors can be made accessible. Weinstein says it is possible, and offers to give a five-lecture course better than the official one, but that almost nobody does it. He compares physics papers to trading Beethoven as sheet music: experts trade scores and nobody actually hears the music.

He pushes back on Fridman, observing that most people are curious about a theory of everything without having digested the physics available since Dirac's 1928 equation. He recalls a New Scientist reporter at his Oxford talk who had never heard of the Dirac equation. He says he does not see why it should be his job to digest ninety years of physics for everyone, though he wants to build a blueprint for the journey.

His simple description of gauge theory is that it is a theory of differentiation where rise over run is measured not from a flat horizontal but from a custom reference level that is internal to the theory. He compares it to Mount Everest's height being quoted "above sea level", when Nepal is landlocked and the real reference is the geoid. He shows with his hand and a phone that the reference level may vary from place to place.

Weinstein closes the earlier thread by describing what he calls the chimeric tangent bundle, an object standing in for the tangent space that nearly has length and angle. Taking unadorned spinors on it gives very large spinor objects on Y14. Pulling this information down to four dimensions, he says, they look like the adorned spinors of ordinary physics, because the ten-dimensional complement, a normal bundle, generates internal quantum numbers such as the charges of up and down quarks.

The chain, as he summarises it, runs from X4 to Y14, to the chimeric tangent bundle, to unadorned spinors, and back down as adorned spinors in the right number. He says the theory yields two copies of a generation of matter plus an "imposter" third one that broke off, a possibility he says regular physics never discussed. In his fermion-and-boson picture, fermions are the artists and bosons are the brushes and paint.

Asked about falsifiable predictions, Weinstein says it predicts that the matter we should see next has particular properties, such as weak hypercharge and response to the strong force. He cannot say at what energy scale it would appear, and says he is not a physicist or quantum field theorist and does not know how that could be detected. He says he hopes there are simple explanations for all of it.

Asked what remains unfinished, Weinstein says he wants parts in better order and that his view differs from the community's. For example, he dates general relativity to the 1913 paper by Einstein and Grossmann, when Einstein saw that the answer would be a curvature tensor on one side and matter and energy on the other. Most physicists date it to the Einstein-Hilbert gradient, and he says he cannot stand that framing.

Fridman, speaking as an enthusiast, says he is confused about whether listeners are meant to follow the technical words or to start a multi-year study, and asks how to play with the ideas and enjoy their beauty. Weinstein replies that he picked one paragraph by Edward Witten, mostly prose, naming three deepest pillars of our knowledge: the Einstein field equation, the Dirac equation and the Yang-Mills-Maxwell equation.

Those three equations appear on a wall in Stony Brook funded by Jim Simons, which in turn needs an owner's manual, Roger Penrose's The Road to Reality. Weinstein calls this the Graph Wall Tome project, running in the Portal community's Discord. He invites listeners to find themselves in that "transmission sequence".

On where Geometric Unity goes next, he says he does not know. He expects the professional community first to get angry and outraged and work through its feelings, then perhaps to come back asking for clarification. So far he has officially seen very little reaction, though it may be happening quietly.