A mysterious substance known as dark matter existed too, its gravity pulling the gas into an elaborate web. As things expanded and cooled, some of the dark matter consolidated in immense orbs, driving the gas to their cores. The rising gravitational pressure within these halos, as astronomers named them, forced hydrogen atoms to fuse into helium, igniting the primordial universe’s first stars.

I watched the spark of cosmic dawn, through 3D glasses. Sitting in front of a projector at the Kavli Institute for Particle Astrophysics and Cosmology at Stanford University, I marveled at filaments of dark matter, a ghostly gray on the screen, branching between halos as the universe stretched. Maelstroms of newly born stars spiraled to the centers of the halos to form the first galaxies.

Scientists have been filling in the universe’s origin story for decades, but in the past year, the largest and most advanced space telescope ever built has rewritten the first chapters. Ancient galaxies glimpsed by the James Webb Space Telescope (JWST) are brighter, more numerous, and more active than anticipated, revealing a frenetic opening to the saga of space and time.

Webb cannot see the first stars, though, as they weren’t bright enough to detect individually. These early monsters blazed hot and grew immense before erupting in supernovae a few million years after flaring to life—a blip in astronomical time.

“We really slowed things down a little bit here,” said Tom Abel, a computational cosmologist and my guide through the simulations. He wore an earring of a human figure curled in the fetal position; it reminded me of the closing shot of 2001: A Space Odyssey, where a child in a womb floats in space. “It’s just so crazy fast. The full realistic version would have been much faster flashes.

Those flashes, the supernovae of stars up to hundreds of times the mass of the sun, transformed the universe. New elements were generated—oxygen to make water, silicon to build planets, phosphorus to power cells—and scattered throughout the expanse. The first stars also broke apart the atoms of the surrounding hydrogen gas, burning away the cosmic haze and making things transparent—a key time known as reionization. As the fog lifted, pockets of stars merged, swirling into bigger and bigger assemblages, including the seed of our own Milky Way.

1.5 billion years shown below

Abel began modeling the birth of the first stars in the 1990s, when no one knew what the earliest astronomical object was, whether a black hole or a Jupiter-size body or something else. Through computer simulations, he and his colleagues helped determine that the first things had to be stars, kindled in places where gravity slowly won out over gas pushing outward. But eventually, Abel moved on from star-birth simulations; he thought there was nothing more to learn.

Then came Webb.

Launched on Christmas morning in 2021, the space telescope is now positioned nearly a million miles from Earth. Its 21-foot-4-inch gold-coated primary mirror captures the light of ancient galaxies, which has been traveling through space for more than 13 billion years, revealing the galaxies as they were in the distant past. (How do you create a telescope unlike anything we’ve had before? These photos show us.)

Astronomers expected to find some of these infant galaxies with Webb. They didn’t expect to find so many—or that the discoveries could shake their understanding of galactic history.

The deepest galaxy survey of the universe ever undertaken kicked off in September 2022, when an international collaboration known as JADES, or the JWST Advanced Deep Extragalactic Survey, began using Webb to observe patches of sky for dozens of hours at a time. Two weeks after observations began, the collaboration gathered in Tucson at the University of Arizona to discuss the first results.

In a modern five-story building with a large, open-air atrium designed to evoke a slot canyon, some 50 astronomers packed into a classroom. A handful stood at the back or brought in extra chairs to sit along the walls. “I’m going to have to start reserving bigger rooms,” said Marcia Rieke, an astronomer at the university and one of the leaders of the collaboration.

The scientists, from tenured professors to twentysomething graduate students, were preoccupied with the mosaic on their laptops: hundreds of images freshly captured by Webb and stitched together. The picture, shared with the team only days before, contained tens of thousands of galaxies and other celestial objects. Excited murmuring ran through the group as they pointed out things to one another that had never been seen: active star-forming regions, glowing galactic centers where black holes might be, and reddish blobs of light from galaxies so distant only Webb could spot them.

“This is a little bit like kids in the candy shop,” Rieke said to me.

Unlike the Hubble Space Telescope, our previous window into the universe’s distant past, Webb was designed to observe in the infrared, which makes it ideal for capturing early starlight. Those rays left their source as ultraviolet but were stretched to redder wavelengths by the expansion of the universe, a phenomenon known as redshift. The higher the redshift, the farther and older the target.

Rieke managed the proceedings with a combination of unfeigned delight and sage reflection, chiming in to answer a technical question or ambling over to a team member between talks to discuss the workings of the telescope. Besides being a lead scientist on JADES, she’s the principal investigator of Webb’s near-infrared camera, or NIRCam—the source of the mosaic of galaxies on everyone’s laptop. She oversaw its design, a 330-pound assemblage of mirrors, lenses, and detectors to drink in the light of the universe and study it through different filters.

“These images are everything we could have hoped for,” she said.

But not everything on the telescope was functioning perfectly. JADES’s near-infrared spectrograph, or NIRSpec, had been experiencing electrical shorts that created spots of light and drowned out astronomical targets in some of the observations. The instrument splits light into spectra, allowing scientists to piece together the chemical composition of a galaxy and precisely measure its redshift. While the NIRCam images could be used to estimate the distances to galaxies, NIRSpec was needed to confirm them.

The electrical shorts delayed some of the team’s observations, a development that turned out to be serendipitous. The astronomers had planned to use NIRSpec to examine objects already known from Hubble, but now they could change the targets to galaxies only just discovered by NIRCam.

“We just went crazy looking through this data that no one had ever seen, looking for these candidates,” Kevin Hainline, an astrophysicist at the University of Arizona, told me later in his office.

One thing the team couldn’t do was change where the telescope was pointing. It had to find objects already in the field of view—and thanks to a bit of luck, four faraway galaxies detected by NIRCam were sitting in the right spot. Two of those, NIRSpec observations would later confirm, were more distant and ancient than any known before.

Two things are already clear about these early galaxies: There are more of them than expected, and they are surprisingly bright for their age. These anomalies could be because the first stars formed more efficiently than thought or there was a larger proportion of big stars than hypothesized. “However star formation gets going in the early universe, it’s not quite like how we might have predicted,” Rieke says.

One early galaxy, GN-z11 from some 440 million years after the big bang, is bright enough that Hubble spotted it in 2016. Now Webb has observed the object as well, including taking its spectrum with NIRSpec.

“This one has everyone sort of confused and excited,” says Emma Curtis-Lake, an astrophysicist at the University of Hertfordshire in England and a member of the NIRSpec team.

Certain elements create bright emission lines in a galaxy’s spectrum, like fingerprints by galactic material. The spectrum of GN-z11 revealed a surprising amount of nitrogen—confounding scientists, who can’t explain its source. Perhaps a population of raging hot stars known as Wolf-Rayet stars scattered nitrogen in pulses of stellar wind. Or maybe several large stars collided, mixing up the material in their cores and surfaces and releasing nitrogen in the process.

GN-z11 may also host a supermassive black hole, which would be remarkable for this early time. It’d be “the most distant black hole that we’ve seen,” Curtis-Lake says.

Obscured at the center of the bright galaxy, it was exposed by spectral lines that Curtis-Lake calls “little hidden monsters.” These lines suggest that material is moving rapidly in a dense area, swirling at roughly a million miles an hour—the kind of thing you would expect to see near a black hole. But how one of these objects could have grown so rapidly remains unsolved.

“This ain’t like it used to be,” said Rieke’s husband, George, as he stepped into a control room that doubled as a kitchenette. “No,” Marcia agreed. “There’s five times as many monitors.”

The couple had offered to show me an old telescope in the mountains near Tucson, where they spent much of their early careers. Both astronomers at the University of Arizona, they’d met in 1972, when George hired Marcia out of grad school. The 61-inch telescope on Mount Bigelow was fairly new then, used to make maps of the lunar surface. It became one of the leading observatories in the budding field of infrared astronomy, a grandfather of sorts to Webb.

The Riekes helped facilitate this succession. While Marcia oversaw the development of NIRCam, George is the lead scientist on Webb’s mid-infrared instrument, or MIRI. The Riekes were trained to stay awake through the night, slowly adjusting the telescope to keep a target in sight as Earth rotated. Their acolytes today can do most of their work from laptops. Just a bunch of wimps,” George quipped.

In the 1970s Marcia and George used the telescope on Mount Bigelow to make some of the first infrared observations of the Milky Way’s center. Scientists had assumed this part of our galaxy was “a collection of old, uninteresting stars,” Marcia explained. But in infrared light, turbulent pockets of gas with rapid star formation were revealed. “That whole picture got changed,” George added.

At the time, the infrared light of the cosmos was only just coming into view. New sensors tuned to the infrared revealed this previously hidden part of the electromagnetic spectrum, which is the full range of light, from gamma rays to radio waves. The telescope on Mount Bigelow helped fill a gap in observations of the local universe, and Webb has similarly plugged a hole in our view of the deep cosmos

Many of the early-career astronomers working on Webb are nearly frantic in their excitement, breathlessly discussing new discoveries and racing to publish scientific papers. Marcia and George, who helped reveal new wonders of our own galaxy, don’t seem to be in that kind of rush. The space observatory is working well, and the cosmic missives it has begun to receive will be deciphered in due time.

But to fully understand our cosmic origins, we will need more than just Webb.

On a recent April morning, I squinted in the sunlight on an expansive plateau between snowcapped volcanoes in Chile’s Atacama Desert. Plastic tubes tickled my nostrils with the flow of oxygen, a requirement for anyone visiting the 16,400-foot-high site of the Atacama Large Millimeter/submillimeter Array (ALMA).

The sky was a deeper shade of blue up there, with fewer molecules in the atmosphere to scatter the light—the very thing that makes this place perfect for astronomy. Towering before me were dozens of four-story-tall radio dishes, white sentinels scattered across the Chajnantor plateau. They pivoted in unison to lock onto a new target.

Among the most advanced radio observatories on the planet, ALMA is also one of the few tools capable of examining the early galaxies being discovered by Webb, albeit in a different light. Webb captures starlight punching through the dust of these galaxies, while ALMA searches for the glow of the dust itself, heated by the stars within.

“These first dust grains come from supernova explosions, so you can indirectly obtain information about the first supernova explosions and the first population of stars,” says María Emilia De Rossi, an astrophysicist at the Institute for Astronomy and Space Physics (IAFE) in Buenos Aires.

ALMA has trained its radio dishes on some of the early galaxies, but in most of its first attempts, the array wasn’t able to find any dust emissions. This could mean that the galaxies are in their infant stages and have not yet produced much dust through stellar explosions, or it could mean that some are actually closer than thought.

In one case, ALMA detected an emission line just beside a target from Webb, perhaps indicating that the galaxy’s stars had blown the dust away or that two galaxies in different phases of their lives were in the process of merging.

ALMA’s first attempts to detect the galaxies discovered by Webb were only glances, short-duration observations slotted into its busy schedule. Astronomers plan to point the array at some of these galaxies for longer periods, searching for faint signals that could reveal how much dust they have generated and, crucially, how many heavy elements they have produced—an indication of how far along they are in galactic evolution.

Toward the end of my visit, I stopped by an enormous hangar at ALMA’s operations facility, lower at 10,000 feet. Two of the towering radio dishes had been brought down from the high site on a 28-wheel transporter vehicle. Workers on lifts were busy replacing some of the dishes’ components, part of a series of upgrades to make the observatory even more capable.

Soon the dishes would be returned to the plateau—ready to swing their gaze back to the firmament, primed to tackle the mysteries of primordial galaxies.