Neil Turok's stunningly simple, testable new theory of the universe

ELI5 / TLDR

For 50 years, physicists have tried to explain the universe by adding stuff: extra particles, extra forces, extra dimensions, parallel universes. None of it has predicted anything that turned out to be true. Neil Turok’s pitch is the opposite. Look at what we actually measure, and the universe turns out to be embarrassingly simple. He thinks that simplicity is a clue, not a coincidence, and he’s built a theory that needs almost no new ingredients. It explains dark matter, says the universe has a mirror twin on the far side of the Big Bang, and makes one sharp prediction that telescopes will confirm or kill within three or four years.

The Full Story

The complaint

Turok starts with a grievance. The most powerful microscope ever built, the Large Hadron Collider, has found nothing that wasn’t already predicted back in the 1960s. The largest telescopes look out at the universe and see, more or less exactly, what people sketched out 50 years ago. And yet theoretical physics has wandered off into a thicket of extra dimensions, multiverses, and undiscovered particles.

“We tried our best but those approaches seem not to have worked… just created more and more complication and zero predictions, zero correct predictions.”

He counts himself among the guilty. He spent years on 11-dimensional M-theory. His conclusion now is blunt: somewhere we took a wrong turn, and the way back is to stop adding things and go back to basics.

His north star is an old line attributed to Michael Faraday, who lectured in the very room Turok is speaking in: there is no law governing any part of the universe that doesn’t also show up in the chemistry of a candle. Think of it as the claim that the same physics runs everywhere, from a flame in a lab to the edge of the observable universe. The big discovery of the last 30 years, Turok says, is that this is literally true. The universe is governed by laws so simple they could have been guessed, and in many cases were.

Five numbers run the cosmos

Here is the headline fact. The standard model of cosmology, the one that fits all the data we have, runs on just five numbers.

Picture yourself at the center of a series of shells. The further out you look, the further back in time you see. The outermost shell you can see is the leftover glow of the Big Bang. Beyond that, a point: everything we see came out of something with zero size.

The five numbers that describe the whole thing: how much dark energy there is (about 70% of everything, a constant Einstein guessed at on the basis of no data at all), how much dark matter there is compared to ordinary matter (a ratio of about five), how many particles there are per photon, and two numbers describing the tiny ripples in the early universe. Those ripples are the seeds that grew into galaxies, and they are about one part in 100,000. That’s it. Five numbers and you fit everything.

Turok offers his own piece of evidence. Years ago, in Princeton, he and colleagues calculated what the polarization of the Big Bang’s afterglow should look like, using equations a physicist named Chandrasekhar wrote down in the 1930s. He went and told the experimentalists to go measure it. When they did, the data fell exactly on the predicted curve, with no knobs left to turn.

“There are no free parameters in that curve… this is an absolute prediction and it fits beautifully.”

His conclusion: the universe is utterly simple, and the laws we learned in laboratories work perfectly.

The mirror universe

Now the strange part. Turok and a collaborator (Latham Boyle) noticed something about how the universe behaves if you trace it back to the Big Bang.

Here’s the idea, gently. The “size” of the universe shrinks as you wind the clock backward, hitting zero at the Big Bang. The question is how it hits zero. Turok found that in the simplest description, it hits zero in the cleanest mathematical way possible, like a straight line touching a point. And because the math behaves so nicely at that point, you can simply continue it through to the other side. Do that, and the universe re-emerges as an exact mirror image of our own behind the Big Bang.

Is the mirror universe real? Turok hedges. Probably not literally. But think of it like a trick from optics. To work out how light bounces off a mirror, you can either do hard equations at the mirror’s surface, or you can just imagine a reflected copy of yourself behind the glass and use the easy free-space equations. Same answer. The mirror universe might just be that trick, a clean way to handle the boundary at the Big Bang.

This connects to a deep symmetry called CPT, which is about the most fundamental symmetry physics has. Roughly, it says the laws of physics shouldn’t care if you flip every charge, mirror-flip space, and run time backward, all at once. The simplest assumption is that the universe doesn’t break this symmetry. The mirror picture is what you get when you take that seriously.

Dark matter, sitting in plain sight for 50 years

This is where the picture earns its keep. The mirror-universe setup, Turok argues, revives an explanation for dark matter that has been available since the 1970s but went unused.

The candidate is something called a right-handed neutrino. The appeal is its blankness: it carries no electric charge, no strong charge, no weak charge. It is, by definition, dark. The only thing it talks to is gravity. That makes it a perfect dark matter candidate, and it’s been known as one for half a century.

So why was it ignored? Because of a habit of thought. Physicists usually figure out how much of a particle should exist by assuming everything started in a hot, mixed-up soup at equilibrium, then cooling off and leaving residues. But a particle that only feels gravity was never in that soup. It broke the recipe. So rather than deal with it, people simply assumed the particle was unstable and decayed away.

“Honestly, literally that’s what they say. Oh, it’s an embarrassment. So we’ll just say it’s unstable.”

Turok’s move: what if it’s stable? Then it’s the dark matter, and you can actually calculate how much of it the Big Bang should produce. The mechanism is the same one that makes black holes glow: a changing spacetime creates particles out of the vacuum (Hawking radiation). Run that calculation through the Big Bang, and the amount of dark matter comes out right if the particle has a specific, very heavy mass.

The one prediction that will be tested

Here is the prediction that matters, and it’s testable within a few years.

If one right-handed neutrino is stable and serves as dark matter, then the lightest of the ordinary neutrinos we already know about must be exactly massless. The logic is a chain of switches: to make the dark matter particle stable, you have to switch off a certain interaction, and switching that one off forces another one off, which forces the lightest neutrino’s mass to zero.

We already know two things about neutrino masses from experiments (the differences between them), but not the absolute scale. Turok’s theory fills in the missing piece by setting the lightest one to zero, which then predicts the total summed mass of all neutrinos.

The way to test it is wonderfully indirect. Neutrinos with mass clump slightly under gravity as the universe expands; massless ones just fly off. So by measuring how galaxies cluster across the sky, you can read off the neutrinos’ total mass. Right now the measurements only set an upper limit. Turok’s theory predicts a specific small value sitting right at the bottom of the allowed range.

“It could be that in three or four years a bump appears at this minimum sum of neutrino masses and that will be incredible confirmation. So we’ll see.”

He’s honest that it could go the other way and rule him out, which he says is fine.

No inflation needed

A big chunk of modern cosmology, the theory called inflation, exists to explain why the universe looks so smooth and uniform in every direction, even between regions that seem too far apart to have ever talked to each other. Inflation says the universe blew up exponentially early on, smoothing everything out.

Turok throws inflation out and explains the smoothness with a different idea borrowed from the physics of gases. Why is the air in a room evenly spread? Not because of some smoothing mechanism, but simply because an even spread is the overwhelmingly most likely arrangement. You don’t need to wait for it to even out; even distribution is just what “typical” looks like when you count the possibilities.

“We don’t need any mechanism. We just need statistical mechanics.”

The bold step is doing this counting for entire universes, not gas molecules. Fortunately the machinery already exists: Stephen Hawking and collaborators built it in Cambridge’s heyday, originally to count the states of black holes. The trick involves making time imaginary, which sounds absurd but is a standard mathematical move that turns a quantum problem into a thermodynamics problem. When you run it, the universe with the most possible states (the most likely universe) is exactly the one we live in: smooth, flat, with a small dose of dark energy.

“Why is the universe the same in two opposite directions? …because it’s correlated, that’s the most likely state. You don’t need any causation.”

The two sides of the sky match not because they ever communicated, but because matching is simply the most likely state. Correlation without causation. That dissolves the puzzle inflation was invented to solve.

Getting rid of the infinities

The last stretch is the most technical, and it’s where Turok thinks gravity itself might turn out simple.

There’s a chronic disease in physics: when you try to combine quantum fields with gravity, the energy of empty space comes out infinite. Gravity is supposed to respond to energy, and here the energy is infinite. You can paper over it with mathematical bookkeeping, but the patches break a deep symmetry (called scale symmetry) that Turok needs for his clean Big Bang.

His fix: look at the sky for a hint. The ripples in the early universe have a particular signature. Ask what kind of field would naturally produce ripples that look exactly like that, and there turns out to be only one mathematical candidate, a peculiar object historically called a “dipole ghost” (first cooked up, his student discovered, by the Indian physicist Homi Bhabha).

The catch with this field is that it lives in a strange space where some quantum states have “negative norm,” which textbooks treat as a death sentence because it sounds like negative probabilities. Turok argues this is a misconception. A negative-norm state is just a label, not a probability. What matters is that the actual probabilities of things happening come out positive and add to one, and he and his students have proven that they do for this theory.

Then the punchline. There is a perfectly respectable, self-consistent theory of quantum gravity, known since the 1970s, that you get by adding certain squared-curvature terms. Physicists rejected it precisely because of those negative-norm “ghosts.” But if the ghosts aren’t actually a problem, the objection evaporates. Turok’s student noticed that a particular limit of this gravity theory is exactly the dipole-ghost field he’d been studying for completely different reasons.

“Maybe gravity is after all rather simple. You’ve got to find some hidden symmetry within quadratic gravity which will eliminate the ghosts… that’s all you have to do, because this theory is just waiting there to be used.”

He calls these baby steps. They don’t yet describe the graviton. But the thread connecting dark matter, the Big Bang’s mirror, the smoothness of the cosmos, and quantum gravity is the same throughout: stop adding, start subtracting, and trust that the universe really is as simple as it looks.

Key Takeaways

The entire observable universe is described by a cosmological model (Lambda-CDM) with just five free parameters. Five numbers fit all the data.
The cosmological constant (dark energy) is ~70% of everything and was guessed by Einstein with no data at all, just on grounds of simplicity. It turned out right.
Particle physics has found no deviations from the Standard Model up to 10 TeV, and cosmology has found nothing not already anticipated 50 years ago. Turok reads this absence of surprises as a clue, not a failure.
The “Hubble tension” and similar anomalies come and go at the 2-sigma level; Turok deliberately ignores them as too weak and too hard to measure to bet on.
Mirror universe: tracing the universe back to the Big Bang, its scale factor hits zero in the simplest possible analytic way, letting you continue the math through to an exact mirror-image universe on the other side. It may be literal, or just a “method of images” trick for handling the boundary condition.
This rests on CPT symmetry (charge-parity-time), the most fundamental known symmetry. The simplest hypothesis is that the universe respects it, rather than breaking it.
Dark matter = a stable right-handed neutrino. It carries no charge of any kind, only couples to gravity, and has been a candidate since the 1970s. It was ignored because it was never in thermal equilibrium, so people just assumed it was unstable to avoid the problem.
Its abundance can be calculated via the gravitational particle-creation mechanism (the same physics as Hawking radiation), and matches observed dark matter if the particle is very heavy.
The testable prediction: if a right-handed neutrino is stable dark matter, the lightest ordinary neutrino must be exactly massless. This predicts a specific minimum value for the summed neutrino mass.
That sum is measured indirectly: massive neutrinos clump under gravity and leave a signature in galaxy clustering. Current data sets only an upper limit; a “bump” at Turok’s predicted minimum within 3-4 years would confirm the theory.
No inflation required. The universe’s smoothness is explained by statistical mechanics, not a smoothing mechanism: a smooth, flat universe is simply the most likely (highest-entropy) state, the way evenly spread gas in a room is.
Matching conditions on opposite sides of the sky need no causal contact: it’s correlation without causation, just the most probable configuration.
The “we can’t talk about anything before the Planck time” rule is, in Turok’s view, an unthinking assumption with loopholes. The cosmos may act as an “ultimate accelerator” letting us see far earlier.
Quantum gravity may already exist: adding squared-curvature (“quadratic gravity”) terms makes gravity renormalizable and asymptotically free. It was rejected only because of negative-norm “ghost” states.
Negative-norm states are not negative probabilities. A quantum state is just a label; what matters is that physical transition probabilities are positive and sum to one (the generalized Born rule on “Krein spaces”).
A limit of quadratic gravity turns out to be exactly the “dipole ghost” / “perfect square” scalar theory, suggesting gravity might be far simpler than string theory or supergravity assume.
Removing the Higgs as fundamental, a peculiar numerology works out: the Standard Model’s particle counts (12 gauge bosons, 48 fermions = 3 generations with right-handed neutrinos) satisfy the anomaly-cancellation equations exactly, where SU(5) and SO(10) grand unified theories fail.

Claude’s Take

This is a serious physicist making a serious, unfashionable bet, and the honesty is the best part. Turok keeps flagging where he’s hand-waving (“we actually don’t know the rules of the game”), tells you his theory could be killed within four years, and admits a prediction of “zero” is the weakest kind of prediction because there are many ways to get zero. That self-skepticism is the opposite of crank energy, and it’s why the talk lands.

The deep appeal of the argument is aesthetic and methodological: the LHC found nothing new, the sky looks exactly as predicted, so maybe the lesson is that we already have most of the laws and the job is to stop bolting on epicycles. That’s a defensible read of the last 50 years. The mirror universe and the gravitational-production dark matter story are elegant and genuinely his (with Latham Boyle), and the one falsifiable handle, the massless lightest neutrino feeding a minimum neutrino-mass sum, is real, near-term, and testable. That alone makes the program worth taking seriously rather than just admiring.

Where to keep a hand on your wallet: the back half leans heavily on rehabilitating “ghost” states in quadratic gravity, and the claim that negative-norm states are harmless is not a settled consensus, however confidently he states it. The Higgs-is-not-fundamental numerology is intriguing but is, by his own admission, numerology, and he holds the Higgs chair while saying there’s no Higgs, which he plays for a laugh but should also make you cautious. None of this is wrong; it’s just early, and “baby steps” is his phrase, not mine. The grade reflects a clear, intellectually honest, idea-dense lecture from a major figure with at least one genuine near-term test. It loses a point because a chunk of the payoff (quantum gravity from ghosts) is promissory, and the broad “everything is simple” framing skates past real anomalies a bit too smoothly.