After a decade of scanning the sky for possible alien technology, UCLA researchers have come back with a very terrestrial answer: more than 105 million candidate signals, and not a single confirmed technosignature. The UCLA Green Bank Telescope SETI campaign sifted through more than 70,000 star systems, only to find that every promising blip eventually traced back to human-made radio interference.
That may sound like a blank result, but in SETI work it is still a useful one. It shows the search methods can survive a brutal amount of noise, and it sharpens the odds for the next round of observations rather than pretending the universe owes us a dramatic reveal.
Green Bank’s 1.15 GHz to 1.73 GHz SETI search
The program used the 100-meter Green Bank radio telescope and listened across a band from 1.15 to 1.73 GHz, a range where narrowband emissions are especially interesting because they can stand out from natural astrophysical chatter. The researchers also tracked Doppler drift, which matters if a transmitter is moving with a planet, and folded in how radio waves can be altered on their trip through interstellar space.
To keep the pipeline honest, the team injected test signals into real observations and checked whether the system could recover them. It did, at rates ranging from 94% to 98.7%, which is a strong sign the search was sensitive enough to catch the kind of transmission it was designed to find. For a field that spends a lot of time staring into static, that is not a bad outcome.
Machine learning did the heavy lifting
Sorting 105.7 million candidates by hand would be a nice way to ruin a decade, so the project leaned on convolutional neural networks and ResNet-style architectures to flag the most interesting traces. Volunteers on Zooniverse also helped build the training sets, which is a reminder that modern astronomy increasingly runs on a combination of algorithms, patience, and unpaid enthusiasm.
- Signals examined: more than 105.7 million
- Star systems surveyed: more than 70,000
- Recovered test-signal rate: 94% to 98.7%
- Telescope: 100-meter Green Bank Radio Telescope
- Frequency range: 1.15 GHz to 1.73 GHz
How rare powerful radio transmitters may be
The most interesting number in the study may be the one that does not sound flashy at all: with 95% confidence, fewer than one in about 16,000 surveyed star systems could host a transmitter above 5×10^16 watts of equivalent isotropically radiated power, at least out to 20,000 light years. That does not rule out intelligent life elsewhere; it just suggests that loud, galaxy-spanning radio beacons are not hanging around every other stellar neighborhood.
There is also a funding story buried inside the astronomy. The authors argue that radio technosignature searches can cover far more space than many biosignature programs, yet public support stays thin, leaving much of SETI dependent on private foundations and education-led efforts. If governments are serious about finding life beyond Earth, they are still paying a suspiciously small bill for the effort.
The next likely step is less about another dramatic ”first contact” headline and more about scaling the tools: better classifiers, broader surveys, and more aggressive use of citizen science. The universe is still silent, but at least the listening equipment is getting harder to fool.