Open questions

Foundations

Ninety-five percent

Here is the inventory of the universe, as measured by the afterglow and confirmed by everything else. Ordinary matter, the stuff of atoms, stars, planets and you: about five percent. Dark matter: about twenty-seven percent. Dark energy: about sixty-eight percent. Ninety-five percent of the universe is something nobody has ever held in a laboratory.

That fact gets misread in two opposite ways. One is to conclude that cosmologists know almost nothing. The other is to hear “dark” and assume it means “unknown in every respect”. Neither is right. Dark means that it neither gives out nor absorbs light. Of both dark ingredients we know the amount to about a percent, we know how they behave, and we know a long list of things they are not. What we do not know is what they are. This lesson is the honest scorecard: what is established, what is active research, and what is open.

Dark matter: what we know

The evidence began in 1933, when Fritz Zwicky measured how fast the galaxies in the Coma cluster were moving and found they should have flown apart long ago unless the cluster held far more mass than its stars. In the 1970s Vera Rubin and Kent Ford measured how fast spiral galaxies rotate and found the same thing: the outer parts spin as fast as the inner parts, which the visible matter cannot explain. The extra mass has since been mapped by the way it bends light passing through it.

The decisive picture is the Bullet Cluster, two clusters of galaxies that collided. The hot gas, which holds most of the ordinary matter, was slowed by the collision and sits in the middle. The mass, mapped by lensing, sailed straight through and sits with the galaxies on either side. Whatever most of the mass is, it is not the gas, and it barely notices a head-on collision.

The afterglow tells the same story from the other end. The pattern of its peaks needs about five times as much matter that feels gravity but not pressure as matter that feels both, and the first minutes independently fix how much ordinary matter there is. Without dark matter, galaxies could not have formed in time. Its existence, as a gravitational fact, is as established as anything in this chapter.

Dark matter: what we do not know

What it is. For forty years the favourite candidate was a heavy particle that barely interacts, a WIMP, and ever more sensitive detectors have been built to catch one: tanks of liquid xenon a kilometre underground, waiting for a nucleus to twitch. They have caught nothing, and have excluded much of the range where a WIMP was expected. Other candidates are very light particles called axions, black holes left over from the first second, and things nobody has thought of. The alternative, that gravity itself works differently at large scales, reproduces the rotation of galaxies but fails the Bullet Cluster and the afterglow. The identity of dark matter is an open question.

Dark energy

In 1998 two teams measured distances to exploding stars halfway across the universe and found them fainter, and therefore further, than they should be. The expansion of the universe is not slowing under gravity, as everyone assumed. It is speeding up, and has been for about six billion years.

Whatever is doing this behaves like an energy belonging to empty space itself, which pushes space apart. Einstein had once written exactly such a term into his equations, the cosmological constant, and then removed it. Put back, it fits every measurement made since. But it comes with the worst embarrassment in physics: quantum theory predicts that empty space should have an energy, and the prediction is larger than the measured value by up to a hundred and twenty powers of ten. Why it is so small, or whether it is exactly constant, nobody knows. The first results from a new survey of galaxies hint, weakly, that it may be changing with time. Established: the acceleration. Open: the cause.

Inflation

The afterglow is the same temperature, to one part in a hundred thousand, on opposite sides of the sky, though in a plain Big Bang those regions could never have exchanged heat. And space is flat to a fraction of a percent, though flatness is unstable and should have been lost long ago. In 1981 Alan Guth proposed that both are explained if the universe grew by an enormous factor in a burst lasting a tiny fraction of the first second. Inflation also predicts the statistical pattern of the afterglow’s ripples, and that prediction has been confirmed in detail.

What it has not delivered is its signature. Inflation should have left gravitational waves that imprint a swirling polarisation on the afterglow. Searches have found none, and the simplest versions of inflation are now excluded. Others predict signals below current reach. Some serious cosmologists argue that a theory flexible enough to fit any result is not really being tested. Inflation is well supported and unconfirmed: active research, with a genuine argument going on.

The tension

The expansion rate today can be inferred from the afterglow: 67.4 kilometres per second per megaparsec, give or take half. It can also be measured directly, from distances to nearby galaxies: 73.0, give or take one. The two have been pulling apart for a decade as the errors shrink, and they now disagree by five standard deviations. Either there is an error nobody has found, or something happened between the afterglow and today that the standard model does not include. Active research, and one of the most important questions in the field.

The beginning itself

Run the equations of general relativity back and they predict infinite density at time zero. Physicists read that as the theory failing, not as a description of an event. Describing the first instant needs a theory that combines gravity with quantum mechanics, and there is not yet one that has been tested. Whether time began, whether anything came before, and whether “before” means anything are open questions.

Smaller open items

Why matter outnumbered antimatter by one part in a billion, from lesson 1.2. Why the oldest stars hold a third of the predicted lithium. What the neutrino masses are: cosmology says the three together weigh less than 0.12 electron volts, particle physics says more than 0.06, and the two are closing in from opposite sides. And whether the universe is infinite: it is flat as far as we can measure, and flat is consistent with both infinite and merely very large.

The point

This is what a healthy science looks like. The model fits thousands of measurements with six numbers, and it says out loud that it does not know what two thirds of its ingredients are. A good map shows where it ends.

Deeper

The evidence for dark matter, line by line

Galaxy rotation curves stay flat to the last measured point, implying mass growing in proportion to radius, in halos extending far beyond the visible disc. The speeds of galaxies in clusters, the temperature of the hot gas held in them, and the lensing of background galaxies through them agree on cluster masses five to ten times the visible. On the largest scales, the afterglow’s third acoustic peak measures the total matter density, while the second peak and the primordial deuterium fix the ordinary matter at about a fifth of it. Structure formation closes the case: the ripples in the afterglow are far too small to have grown into galaxies by today unless a component that does not interact with light began collapsing before recombination.

Candidates and searches

WIMPs would occasionally scatter off nuclei; the LZ experiment’s first run, with seven tonnes of liquid xenon, set limits on the scattering rate that exclude much of the parameter space favoured in the 2000s. Axions would convert to photons in strong magnetic fields; resonant cavity experiments scan the possible masses one narrow band at a time. Primordial black holes are constrained by lensing surveys and by the afterglow to a few remaining mass windows. Collider searches at the Large Hadron Collider look for missing energy. None has produced a signal. Modified Newtonian dynamics, proposed in 1983, fits rotation curves with one new constant and no dark matter, and it is worth knowing why it is not the answer: it underpredicts cluster masses by a factor of a few, cannot produce the Bullet Cluster’s separation, and cannot reproduce the afterglow’s peaks.

Dark energy in numbers

The equation of state, the ratio of dark energy’s pressure to its density, is measured at −1.03 ± 0.03 by combining the afterglow, supernovae and galaxy surveys; a cosmological constant has exactly −1. The DESI survey’s first-year baryon acoustic oscillation results, combined with supernovae and the afterglow, prefer a value that was different in the past, at a significance between about 2.5 and 4 standard deviations depending on which supernova sample is used. That is interesting and not decisive. Weinberg’s 1989 review remains the clearest statement of the underlying problem: any known contribution to the vacuum energy is vastly larger than the total, and they would have to cancel to more than a hundred decimal places.

Inflation’s ledger

Predictions confirmed: a nearly flat universe; ripples that are nearly the same on every scale, with a slight tilt, measured at 0.965 ± 0.004; ripples that are Gaussian and adiabatic. Prediction outstanding: primordial gravitational waves, parameterised by the tensor-to-scalar ratio, now limited below 0.036. The simplest single-field models with a quadratic potential predicted about 0.13 and are excluded; plateau-type models predict values of a few thousandths, within reach of the next generation of experiments. The critics’ argument is that inflation, given freedom over its potential, can accommodate almost any outcome, and that the theory’s own logic leads to an eternal multiverse in which predictions become ambiguous. Both sides publish in the same journals.

The tension in numbers

The afterglow route: fit the six numbers, then compute the expansion rate today. The local route: calibrate Cepheid variables with parallax, use them to calibrate supernovae, use supernovae to measure the expansion. The two disagree by about 8 percent with combined errors near 1.5 percent. Proposed systematic errors on the local side have been tested one by one, most recently with Webb, without closing the gap. Proposed new physics includes an extra dose of dark energy just before recombination, which shrinks the sound horizon and raises the inferred rate; it fits the tension at the price of fitting other data worse.

Neutrinos, curvature, and the size of everything

Cosmology weighs neutrinos by their effect on the growth of structure; the Planck data with galaxy surveys give a summed mass below 0.12 electron volts, and the newest surveys push lower. Oscillation experiments set the floor at 0.06. A laboratory measurement would be a direct test of the whole framework. Curvature is measured at 0.001 ± 0.002: flat to the precision available, which is compatible with an infinite universe and with a finite one at least several hundred times larger than the observable part. The topology, whether space wraps around on itself on some enormous scale, has been searched for in the afterglow and not found.

Frontier

What would settle each

Dark matter: a particle detection, in a xenon tank, an axion cavity or a collider, or a definitive exclusion that forces the field toward the remaining candidates. Dark energy: the completed DESI survey, together with the Euclid space telescope and the Rubin Observatory’s decade-long survey, will measure whether the acceleration has changed with time to a precision that either confirms the hint or kills it. Inflation: a detection of B-mode polarisation, or a limit low enough to rule out the plateau models as well. The Hubble tension: an independent local measurement that does not use Cepheids, from gravitational waves or from the tip of the red-giant branch, at one-percent precision. Neutrino masses: laboratory experiments closing in from above while cosmology closes in from below. The beginning: a testable theory of quantum gravity, which no one yet has.

Reading a list of open questions

Every one of the questions above was once unaskable. That the universe has an inventory at all, that its expansion rate can be measured two ways precisely enough to disagree, that a survey can hint at the behaviour of empty space six billion years ago: these are the achievements that make the questions possible. The next chapter goes back to the first stars and follows what they made. The rest of the story is built from the five percent.

Transcript and shot list

Target runtime about 8 minutes at a measured narration pace, with room for the visual beats to breathe. Shot types: ANIM is rendered from code with Manim, AI is a generated shot from Higgsfield with the shared style reference, REAL is agency imagery with credit, CAM is the narrator. Timecodes are targets and will move to match the recorded narration.

[00:00] S01 · AI · A vast dark hall; a single lamp lights one small corner; the rest is shape without detail.

Here is the inventory of the universe. Ordinary matter, the stuff of atoms and stars and you: about five percent. Dark matter: twenty-seven. Dark energy: sixty-eight. Ninety-five percent of everything is something nobody has ever held in a laboratory. Does that mean cosmologists know nothing? No. It means something more interesting, and this lesson is about being honest about exactly what.

[00:40] S02 · ANIM · Cosmic calendar; a marker at late July, labelled “dark energy takes over”; a second marker before the first tick, labelled “dark matter, from the start”.

Dark matter has shaped the universe since before the first tick of the calendar. Dark energy took over the expansion around late July, six billion years ago. Both are still at work tonight.

[01:00] S03 · ANIM · A spiral galaxy; the expected rotation curve falls away with radius; the measured one stays flat; the gap is shaded.

Start with what we know about dark matter, because it’s a lot. In 1933 Fritz Zwicky found that galaxies in a cluster move far too fast to be held by their stars. In the 1970s Vera Rubin measured how galaxies spin and found the outer parts moving as fast as the inner parts, which the visible matter can’t do. Something unseen, five times heavier than everything we can see, holds galaxies together.

[01:45] S04 · REAL · The Bullet Cluster: X-ray gas in pink, lensing mass in blue (NASA, CXC, ESO, STScI; confirm credit line).

The clinching picture is this. Two clusters of galaxies collided. The hot gas, which holds most of the ordinary matter, was slowed by the crash and sits in the middle, in pink. The mass, mapped by the way it bends light, sailed straight through and sits on either side, in blue. Whatever most of the mass is, it isn’t the gas, and it barely notices a head-on collision. The afterglow agrees: its pattern of peaks needs five parts of matter that feels gravity but not pressure for every one that feels both. Without it, galaxies could never have formed in time. That dark matter exists, as a gravitational fact: established.

[02:40] S05 · ANIM · A tank of liquid xenon a kilometre underground; a counter of years running; no detection; a list of candidates, each with a question mark.

What it is: open. For forty years the favourite was a heavy particle that barely interacts, and tanks of liquid xenon have waited underground for one to bump a nucleus. They’ve caught nothing, and ruled out most of where it was expected to be. Maybe it’s a very light particle called an axion. Maybe black holes from the first second. Maybe gravity itself behaves differently at large scales, though that idea fails the Bullet Cluster. Nobody knows. That’s not embarrassment. That’s the frontier.

[03:25] S06 · ANIM · A plot of supernova brightness against distance; the points bend upward away from the “slowing” curve; a label: accelerating.

Now dark energy. In 1998 two teams measured distances to exploding stars halfway across the universe and found them fainter than they should be: further away. The expansion isn’t slowing under gravity. It’s speeding up, and has been for six billion years.

[03:55] S07 · ANIM · Einstein’s equation with the Lambda term crossed out, then restored; a bar labelled “predicted” towering 120 powers of ten over “measured”.

Whatever is doing that behaves like an energy of empty space itself. Einstein once wrote exactly such a term into his equations and then removed it. Put back, it fits everything. But here’s the worst prediction in physics: quantum theory says empty space should have an energy, and the prediction is bigger than the measurement by up to a hundred and twenty powers of ten. Why it’s so small, nobody knows. And a new galaxy survey hints, weakly, that it may be changing over time. The acceleration: established. The cause: open.

[04:45] S08 · ANIM · Two patches on opposite sides of the sky at the same temperature; the universe balloons in a flash; the ripple spectrum’s tilt matches; then a swirl pattern labelled “not found”.

Inflation. Opposite sides of the sky show the afterglow at the same temperature to one part in a hundred thousand, though they could never have exchanged heat. And space is flat, though flatness should have been lost long ago. In 1981 Alan Guth proposed that both are explained if the universe grew by an enormous factor in a burst inside the first second. Inflation also predicted the pattern of the ripples, and that’s been confirmed in detail. What it hasn’t delivered is its signature: gravitational waves that would leave a swirl in the afterglow’s polarisation. Not found. The simplest versions are already ruled out. Some serious people argue the idea is too flexible to test at all. Well supported, unconfirmed, genuinely argued over: active research.

[05:45] S09 · ANIM · A gauge with two needles: 67.4 from the afterglow, 73.0 from nearby galaxies; the gap labelled five sigma.

Then the tension. The expansion rate inferred from the afterglow: 67.4. Measured directly from nearby galaxies: 73. The two have pulled apart for a decade as the errors shrank, and they now disagree by five standard deviations. Either there’s an error nobody has found, or something happened between the afterglow and today that the model doesn’t include.

[06:15] S10 · ANIM · The equations run backward to time zero and break; a question mark before the first tick.

And the beginning itself. Run the equations back to time zero and they give infinite density, which is the theory failing, not a description of anything. Describing the first instant needs a theory that joins gravity to quantum mechanics, and nobody has one that’s been tested. Whether time began, whether anything came before: open.

[06:45] S11 · ANIM · The full scorecard: a column of ESTABLISHED, a column of ACTIVE RESEARCH, a column of OPEN QUESTION, each entry sliding into place.

The scorecard for the whole chapter. Established: the universe was hot, dense and small, it’s expanded and cooled for 13.8 billion years, its recipe, its afterglow, its inventory, that dark matter exists, that the expansion is accelerating. Active research: what inflation was, the tension, whether dark energy changes, the first stars. Open: what dark matter is, what dark energy is, why there’s more matter than antimatter, what happened at the start. A good map shows where it ends.

[07:30] S12 · AI into ANIM · The dark hall again; the lit corner is small but everything in it is sharp and detailed; the calendar returns.

So, ninety-five percent dark. Here’s the right way to hear that. Every one of these questions was once unaskable. That the universe has an inventory at all, that we can measure its expansion two ways precisely enough to disagree, that we can ask what empty space was doing six billion years ago: those are the achievements that make the questions possible. Next chapter, we go back to the first stars and follow what they made. The rest of the story is built from the five percent.

Shot list

Two AI shots, nine code-rendered shots, one agency image. Estimated Higgsfield use: 3 to 6 generations including retries.