Posted by There is no agency behind the SETI program, no headquarters with a plaque on the door. What exists is a scattered handful of institutions – a nonprofit in Silicon Valley, a research center at Berkeley, a ten-year grant from a venture capitalist – that between them keep a few dozen people employed full time on the question of whether anybody else is transmitting. The whole field costs less per year than a mid-sized planetary mission.
That is not an accident of scale. It is the direct result of a decision made in Washington in 1993, and of a small number of very wealthy people who decided afterwards to keep the lights on anyway. Understanding the SETI program today means understanding its plumbing: who signs the cheques, who owns the dishes, and who gets to decide where those dishes point on a given Tuesday night.
The Year Congress Turned Off the Money
NASA once ran a genuine, funded search. The High Resolution Microwave Survey switched on in October 1992, timed to the 500th anniversary of Columbus making landfall, and it had two halves: a targeted survey of nearby sun-like stars using Arecibo, and an all-sky sweep using the Deep Space Network antennas at Goldstone. It lasted a year. Senator Richard Bryan of Nevada stripped the funding in 1993 and announced, with some satisfaction, that the Great Martian Chase was over.
The SETI Institute, founded in Mountain View in 1984, picked up the wreckage of the targeted half and relaunched it under a name chosen with no subtlety whatsoever: Project Phoenix. From 1995 to 2004 it worked through roughly 800 stars within about 200 light years, borrowing time on Parkes in Australia, the 140-foot dish at Green Bank, and Arecibo. Every hour of it was paid for privately. The donor list reads like a founding charter of the American tech industry – Paul Allen, William Hewlett, David Packard, Gordon Moore, and Barney Oliver, Hewlett-Packard’s former research chief.
Phoenix found nothing. Every candidate that looked promising for a few minutes turned out to be us.
Forty-Two Dishes in a Cow Pasture
Borrowed telescope time has an obvious weakness, which is that you have to give it back. The answer, worked out jointly by the SETI Institute and the Radio Astronomy Laboratory at UC Berkeley, was to build something nobody else needed to share. The Allen Telescope Array sits at Hat Creek Radio Observatory in the volcanic country of northern California, a few hours northeast of San Francisco, and it took the opposite approach to almost every radio observatory before it: a swarm of small cheap dishes instead of one enormous expensive one, with the computing doing the hard part. Six-metre dishes, off-the-shelf satellite-television parts where possible, a wide field of view, several stars at once.
The design called for 350 of them. Paul Allen put in more than 25 million dollars, the array carries his name, and 42 dishes were built. That is where it stopped.
In April 2011 the array went into hibernation – literally powered down – when Berkeley’s funding evaporated during the California budget crisis. It came back that December on a mix of public donations from a small crowdfunding campaign and a US Air Force contract to track orbital debris, which is the kind of compromise a field makes when it is out of money. Berkeley withdrew entirely in 2012.
Since 2019 the array has been rebuilt from the feeds outward, funded by Franklin Antonio, a co-founder of Qualcomm, who died in 2022 and left the work provided for. The new receivers cover a continuous band from roughly 1 to 12 gigahertz where the old ones covered patches; the digital back end was replaced wholesale. It matters because the Allen Telescope Array remains the only instrument on Earth whose primary job description is listening for technosignatures. Nobody has to approve the observing plan. The institute owns the dishes.
The Screensaver Years
Meanwhile Berkeley had been running the other half of the strategy: take the data somebody else is already collecting, and process it for free.
SERENDIP rode along on Arecibo for years, recording whatever the dish happened to be pointed at while other astronomers worked. On 17 May 1999 the Space Sciences Laboratory launched SETI@home, built by Dan Werthimer and David Anderson, which chopped that piggybacked Arecibo data into small work units and shipped them to anyone willing to install a screensaver. Millions of people in more than 200 countries did. At its peak the project commanded more raw computing throughput than any single supercomputer of the era, assembled entirely out of idle office PCs.
On 31 March 2020 it stopped sending out new work. Not a collapse – the team’s explanation was mundane and honest: they had processed enough of the back catalogue, and the bottleneck had moved from raw cycles to the slow scientific work of sifting the results. Arecibo’s own collapse eight months later, on 1 December 2020, settled the question of fresh data regardless. The lasting product of SETI@home was not a signal but BOINC, the volunteer computing platform spun out of it in 2002, which now runs protein folding and climate models for people who have never thought about aliens at all.
What remained at Berkeley consolidated into the Berkeley SETI Research Center under Andrew Siemion, and that is where the next thing landed.
What a Hundred Million Dollars Buys
In July 2015, at the Royal Society in London, with Stephen Hawking sitting beside him, Yuri Milner announced Breakthrough Listen: 100 million dollars over ten years, the largest sum ever committed to the search. The stated targets were the million nearest stars, the plane of the Milky Way, and a hundred nearby galaxies.
The revealing detail is what the money mostly buys. Not new telescopes – time on existing ones. Listen purchased a substantial share of the Green Bank Telescope’s schedule in West Virginia and of Parkes in New South Wales, plus an optical laser search on the Automated Planet Finder at Lick Observatory. The rent arrived at Green Bank at a moment when the National Science Foundation was looking to reduce what it spent on the site, which tells you something about American radio astronomy right now.
Then came the smarter arrangement. Since 2018 Listen has been observing commensally with MeerKAT in South Africa, tapping the data stream from its 64 dishes while other projects direct the telescope. No competition for time, millions of stars surveyed as a by-product of somebody else’s science.
Listen also releases its data publicly, in petabytes, which is how the field’s most famous recent embarrassment got resolved properly. In archived Parkes data from April and May 2019, pointed at Proxima Centauri, an analyst found a narrowband signal near 982 megahertz drifting the way something on a moving planet should. BLC1 leaked to the press in December 2020; two Nature Astronomy papers in late 2021 concluded it was terrestrial interference, clinched by a family of look-alike signals in the same data that could only have come from local electronics.
It was a transmitter. Just not a distant one.
Who Decides Where the Dish Points
Ordinary astronomy allocates telescope time by proposal. You write a case, a time allocation committee ranks it against everyone else’s, and the good instruments are oversubscribed several times over. SETI proposals have historically done badly in that process for a reason that is hard to argue with: you cannot promise a result, and a committee choosing between a guaranteed measurement and a maybe picks the measurement.
So the field stopped asking. The three surviving strategies: own the telescope, buy the hours, or ride along invisibly. That last one is now the default. COSMIC, a system built by the SETI Institute, has been taking a copy of the data from all 27 antennas of the Very Large Array since 2023, processing it for technosignatures while the array carries out its own sky survey. Millions of pointings, not one dedicated night requested. China’s 500-metre FAST telescope hosts a similar back end through a Berkeley collaboration, which produced candidate signals in 2022 that turned out to be interference. When the Square Kilometre Array comes online, the plan is the same.
The Part That Keeps Breaking
Ten-year grants end. Milner’s commitment ran from 2015, and nothing has been announced about what replaces it. NASA has crept back in cautiously – a technosignature workshop in Houston in 2018, then a small grant in 2020 to a team led by Adam Frank at Rochester to study technosignatures beyond the radio band, the agency’s first money in the area in decades. Small sums. Larger symbolism.
The real vulnerability of the SETI program is that its funding is personal rather than institutional. Allen, Antonio, Milner: three people, three cheques, and an entire field’s observing capacity resting on their continued interest. Donors die. Priorities move. The instruments are mortal too – Arecibo fell down, and the Big Ear dish in Ohio that recorded the Wow! signal in 1977 was demolished in 1998 to extend a golf course.
Jill Tarter, who spent decades as the public face of the search and was the model for the astronomer in Contact, stepped away from day-to-day research in 2012 to raise money full time. She was explicit about why. The physics was not the constraint. The chequebook was.
If you want to follow where the next hundred million comes from, and which dish gets pointed where, SETIworld tracks the search as it happens – the observing runs, the candidate signals that do not survive follow-up, and the ones nobody has explained yet. Come and read along.
