The first alien will not speak. It will be a pattern too faint for human eyes, caught by a machine built to erase a sun.
A few missing colours in a distant planet's light will reveal an atmosphere. That atmosphere will contain gases that should destroy one another, yet remain in stubborn abundance. The signal will return when the planet is observed again. It will strengthen when another telescope looks. Every lifeless explanation will fail.
Then the world will change.
ParallaxSee forecasts that by 2060, artificial intelligence will identify the first broadly accepted evidence of life beyond Earth in the atmosphere or on the surface of a rocky exoplanet. House confidence: 76%.
The discovery will not resemble a spaceship landing or a radio voice crossing the dark. It will be an argument assembled from photons: a planetary spectrum, repeated over years, searched by AI across an immense library of possible atmospheres and finally made too coherent for geology alone to explain.
The decisive instrument is already taking shape. NASA's Habitable Worlds Observatory is being designed to suppress the light of nearby stars and directly examine roughly two dozen potentially habitable planets. Its preliminary catalogue contains about 160 nearby stars. By 2060, that narrow but profound census will have accumulated something astronomy has never possessed before: repeated, comparable spectra of other possible Earths.
01 — The telescope will be built to erase a sun.
Earth is easy to miss. Seen from another planetary system, it would be a small blue-grey point approximately 10 billion times fainter than the Sun in visible light. Trying to photograph it is like searching for a firefly pressed against a lighthouse from thousands of kilometres away. NASA's exoplanet-imaging technology programme
The Habitable Worlds Observatory, or HWO, is intended to win that contest. NASA's present concept is a roughly six-metre space telescope observing ultraviolet, visible and near-infrared light. Its defining instrument will be a coronagraph: an optical system that blocks the star while allowing the faint light beside it to survive. NASA's Habitable Worlds Observatory
This is more difficult than placing a black disc over the middle of a photograph. Tiny imperfections in a mirror scatter starlight across the image. A shift smaller than an atom can create a false point beside the star or bury a real planet beneath a bright optical speckle. HWO therefore requires deformable mirrors that change shape with extraordinary precision, flattening the incoming wavefront and carving a dark zone around the star. NASA's test programmes are working towards mirror stability measured in tens of picometres and starlight suppression approaching one part in 10 billion. NASA's deformable-mirror programme, NASA's ultra-stable observatory testbed
That is what makes this telescope special. Hubble showed us galaxies. Webb is opening the infrared universe and probing selected exoplanet atmospheres. HWO will be designed around a still more audacious task: remove a nearby sun from the picture so that an Earth can appear.
02 — Its coverage will be a census, not a panorama.
No telescope in the 2040s will inspect every star in the Milky Way. It will not need to.
NASA's preliminary HWO target list contains roughly 160 nearby stars accessible to a six-metre observatory. From that neighbourhood, the mission is being designed to directly image about 25 potentially habitable worlds and examine their atmospheres. The working launch ambition is the first half of the 2040s. NASA Exoplanet Exploration Program target-list announcement, HWO's science objectives
Twenty-five may sound small. In planetary science it is a revolution. Today, researchers often struggle to interpret one faint spectrum from one unusual world. HWO will turn the search into a comparative science. A candidate Earth will sit beside hotter planets, colder planets, dry planets and clouded planets observed with the same instrument. Researchers will be able to ask not merely whether one planet contains an interesting gas, but whether its entire atmosphere is exceptional among its neighbours.
By 2060, the important form of coverage will be time. Promising worlds will have been revisited at different points in their orbits. Their atmospheres will have been tested across seasons, changing cloud cover and stellar activity. Candidate signals will have been compared with observations from giant ground telescopes and successor instruments.
The telescope will not sweep the whole sky. It will do something scientifically stronger: return to the nearest possible Earths until their skies become familiar.
03 — A spectrum is an atmosphere turned into a barcode.
A planet is too distant to reveal a forest, an ocean wave or an animal. But its light carries a chemical inventory.
HWO will separate the planet's reflected light into wavelengths. Molecules absorb particular colours, leaving dark lines and bands in the spectrum. Oxygen, ozone, water vapour, methane and carbon dioxide each alter that barcode differently. Temperature, pressure, clouds and the planet's surface alter it again.
The discovery will not depend on finding one fashionable molecule. Oxygen can be produced without life. Methane can emerge from geology. Water proves habitability, not habitation. The compelling signal is a system of chemicals and conditions that should not persist together unless something continually replenishes them. Life on Earth has transformed the atmosphere into such a state of chemical disequilibrium. A living exoplanet should leave its own version of the same contradiction.
NASA's stated HWO strategy reflects this. The observatory is intended to measure possible biosignature gases while also establishing the planet's broader environment, precisely because context is needed to distinguish biology from a false positive. HWO's science objectives
The extraterrestrial signal may be unfamiliar. It may not reproduce Earth's oxygen-rich history. What matters is not that the barcode looks like home. What matters is that its pieces form a planetary metabolism: energy enters, matter cycles and the atmosphere remains far from the state an empty world should settle into.
04 — One dramatic molecule will not be the discovery.
The path to 2060 is already teaching astronomers what not to accept.
The atmosphere of the exoplanet K2-18 b produced one of the most publicised biosignature debates of the Webb era. Yet when researchers expanded the range of atmospheric models, the apparent evidence for particular gases weakened or disappeared. Low signal, incomplete chemistry and model choices could imitate confidence. Nature Astronomy's analysis of the K2-18 b claims
This is not a reason to expect failure. It is the rehearsal that makes success credible.
By the time HWO begins its mature survey, astronomers will know that a single best-fitting model is not enough. The life-bearing explanation will have to survive different cloud assumptions, temperatures, chemical networks, stellar histories and data-processing methods. It will have to predict later observations before they arrive.
A genuine biosphere will keep winning those tests. Noise will move. Instrumental artefacts will change with the telescope. A mistaken chemical model will collapse when another wavelength is measured. But a planet's atmosphere will still be there on the next orbit, maintaining the same connected pattern.
Alien life will be announced only after scientists have tried very hard to kill the claim. By 2060, one claim will refuse to die.
05 — AI will search the planet that no human can see.
The telescope supplies the photons. AI makes the search exhaustive.
First, machine-learning systems will help separate a planet from the changing optical speckles around its star. They will learn the behaviour of the telescope, identify patterns that move like an orbiting world and direct follow-up observations towards the most valuable targets.
Second, AI will compare each faint spectrum with molecular templates. High-resolution spectroscopy can search for the repeated forest of absorption lines made by a gas even when no individual line is obvious. Cross-correlation aligns those tiny features like hundreds of weak fingerprints. Modern reviews describe this method as a route from detecting isolated molecules towards characterising whole atmospheres. Nature Reviews Physics on high-resolution exoplanet spectroscopy
Third, AI will perform atmospheric retrieval. Researchers will generate enormous libraries of synthetic planets by varying gases, clouds, temperatures, surfaces and stellar radiation. A model trained on those simulated spectra can rapidly estimate which combinations fit the real light and how uncertain each answer remains. Machine-learning retrieval has already reproduced full probability distributions for molecular abundances and clouds in exoplanet spectra; newer work connects the training data to three-dimensional climate simulations. Machine-learning atmospheric retrieval in Nature Astronomy, NASA's Bayesian machine-learning framework
Finally, AI will look for the relationship humans did not think to request. It can compare one planet with the others around the same star, test whether an atmospheric combination is rare across the survey, and notice a recurring anomaly spread across many wavelengths and years.
This is where the discovery becomes likely. A scientist can inspect a spectrum. An AI system can test that spectrum against a library of possible worlds, reject millions of near-matches and return the small family of explanations that survives. It will not declare that aliens exist. It will reveal that every non-living model is losing.
06 — The strongest biosignature will be comparative.
The most persuasive evidence may come not from what a planet possesses, but from what it lacks.
One proposed strategy is to compare rocky planets in the same system. A world with oceans and biology may have substantially less atmospheric carbon than its otherwise similar neighbours because carbon has been dissolved, buried or incorporated into biomass. Researchers have proposed atmospheric carbon depletion as a practical sign of liquid water and potentially life, especially when several sibling planets provide a natural control experiment. Nature Astronomy on atmospheric carbon depletion
This is exactly the kind of reasoning an AI-assisted survey can amplify. Instead of asking whether carbon dioxide on one distant dot is high or low, the system can compare entire planetary families. Instead of treating clouds as an inconvenience, it can search for recurring differences between cloudy, dry and ocean-bearing worlds. Instead of fitting one static atmosphere, it can track which gases rise and fall together through a planet's year.
The first detection of life will therefore be less like recognising a face and more like solving a case. One spectrum provides the clue. Other planets provide the controls. Climate models reconstruct the scene. Repeated observations test the alibi.
AI will hold all of those relationships at once. That ability—not synthetic eloquence—will make it indispensable.
07 — Other observatories will turn a finding into a discovery.
HWO will not work alone.
NASA's Roman Space Telescope is already intended to demonstrate the high-order wavefront control and coronagraphic techniques required to suppress starlight, creating a technological bridge towards HWO. NASA on Roman's coronagraph
On Earth, the Extremely Large Telescope is scheduled to begin scientific work around 2030. Its enormous mirror and high-resolution spectrographs will study nearby planetary systems with methods different from a space coronagraph. European Southern Observatory's ELT timeline
By 2060, a promising HWO world can therefore be attacked from several directions. One instrument will directly image it. Another will search for individual molecular line patterns. Observations taken years apart will test whether the signal follows the planet rather than the telescope or its star. Independent teams will rebuild the atmospheric model with different assumptions.
This redundancy is the point. The world will not accept alien life because one AI assigns a high probability to one spectrum. It will accept alien life when several instruments keep finding the same planetary chemistry and every serious abiotic explanation requires increasingly impossible coincidences.
The AI will locate the pattern. The observatories will make it repeatable. Scientists will make it a discovery.
08 — By 2060, the alien sky will be harder to explain without life.
The search for life has spent centuries moving its target away from fantasy and towards measurement. Mars became terrain. Exoplanets became points in a catalogue. Their atmospheres are now becoming data. The next step is comparison.
HWO's deepest achievement will not be one beautiful photograph. It will be a disciplined collection of possible Earths observed by the same stable instrument. AI will turn that collection into a map of what dead planets normally look like. Against that background, a living world will stand out.
The first biosignature may be oxygen maintained beside a reducing gas. It may be a carbon pattern shared by an ocean and its atmosphere. It may be a seasonal surface signal, or a combination no present textbook contains. The exact chemistry is uncertain. The arrival of a searchable, repeated planetary census is not.
By 2060, one nearby rocky world will have produced the same extraordinary answer through multiple observations: its atmosphere is performing work that sunlight, volcanoes, oceans and rocks cannot sustain by themselves.
We will not see the organisms. We may not know whether they are cells beneath an ocean, a film across wet ground or something with no earthly name. Yet humanity will know that biology happened twice.
The first alien will be a stubborn chemistry in a distant sky. A telescope will rescue it from the glare of its sun. AI will recognise the pattern—and, by 2060, the universe will no longer look sterile.
Loading interventions…