 On the surface of an average planet, circling an ordinary yellow star, an advanced intelligence searches the skies for evidence of life. Directed by even higher intelligence, machines with brains of silicon patiently sift through faint shards of radio data for the unmistakable signal that will indicate the first sign of life beyond Earth. In 1959, physicist Philip Morrison was co-author of the first scientific paper to suggest a strategy for such a search. What I would like to know is an answer to a very simple question. Are we alone, as conscious beings in this entire buzzing 400 billion star galaxy, one of ten to the tenth other galaxies? It seems pretty implausible. It's an enormous task to search the skies for intelligent life, looking for a golden needle in a huge cosmic haystack, but SETI, the search for extraterrestrial intelligence, has come a long way since early experiments by young radio astronomer Frank Drake. In 1960, Drake made the first radio search from Green Bank, West Virginia. He called it OSMA, after a princess from the fictional land of Oz. Now plans for the most sophisticated SETI search ever focus on the Goldstone Deep Space Communication Complex in California's Mojave Desert. It is the site of tests for a possible future NASA project, one not yet funded. The full-scale NASA system, when operational in the next decade, would be billions of times more powerful than the sum of all previous searches. For Carl Sagan, a proponent of SETI for many years, this technical progress has made the present unique. For the first time, we're mustering substantial, sophisticated, serious scientific searches for extraterrestrial intelligence. There's never been a time like that before. So there is some chance that in the next few decades, we will get a signal from some spectacularly distant, spectacularly exotic civilization, and everything on Earth will as a consequence change. That is possible. Why, when, and whether to search for life beyond Earth has been debated for centuries. Deciding what sort of signal to look for in the skies is no easy task. NASA's proposed search focuses on radio, a portion of the electromagnetic spectrum where nature produces the least interference for any intelligent signal. The plan is to use existing radio antennas and combine them with advanced computer hardware and software, specifically designed for the task. Signal processing equipment suitable for SETI is constantly becoming more powerful and more efficient, but also cheaper and smaller than ever before. In the mid-1990s, NASA hopes to deploy its systems in a relatively low-cost 10-year search that will, for the first time, systematically explore all the radio frequencies and directions on the sky that researchers think might reveal a signal. My name is Jill Tarder. I'm an astronomer from the University of California at Berkeley, and I work with colleagues from NASA's Ames Research Center, the Jet Propulsion Laboratory, and the SETI Institute to help design the special-purpose tools needed to do SETI properly. Over the years, I've worked at many radio telescopes, like this one at Nassay in France, trying to adapt existing equipment to this difficult task. These efforts, I hope, will soon be eclipsed by the planned NASA system, some parts of which may eventually be placed here at Nassay. Contemporary SETI is a science, a discipline which transforms humanity's age-old speculations about life in the universe into experimental reality. SETI pioneer Frank Drake came up with a way to organize our developing knowledge and current ignorance. SETI scientists often use what's called the Drake equation to illuminate the necessary conditions for contact and to provide a rough estimate of the number of other civilizations. The existence of other technical civilizations depends upon astronomy, how planets form, comparative planetology, biochemistry, the role of intelligence in evolution, technology, and the fate of technical civilizations. So SETI becomes a way to test our theories of the origin and evolution of the universe and the place of life within it. Some factors in the Drake equation are well-determined, others are not. We're pretty sure our galaxy has about 400 billion stars and maybe 10% of them will shine long enough for life to evolve and we understand something of the life cycle of those stars. We know that it required colossal supernova explosions to transform the simple elements produced in the Big Bang into the heavy elements that form rocky planets and we ourselves. And we have taken the first steps in understanding the chemical origin of life and the principles of evolution. But as yet we found no definitive proof of solar systems beyond our own, though recent observations of the disk of matter around the nearby star Beta Pictoris suggest the planets may be common. In our own solar system we have found no signs of advanced life. Though planets like Mars and the large moons like Europa and Titan may have provided suitable environments for life in the past or may do in the future and seem worth further exploratory missions. So far we have detected no signals from other civilizations around other stars using technologies which could make them visible across the light years. And we cannot predict whether civilizations with advanced technology will perish quickly or learn to use it to explore the universe and thrive for the lifetime of their star. Some people look at the seven conditions Drake proposed and see a weak chain that could easily be broken. Drake himself thinks there could be many thousands of civilizations with whom we could communicate in the Milky Way alone. We cannot take the solar system which we know has happened and the life on Earth as typical. And as far as we know it is typical. We know of nothing, no freakish event that was required for us with our motorcycles and our videotape recorders to exist. SETI researchers such as Harvard's Paul Horowitz doubt that the question of other life in the universe will ever be answered in the abstract. People have argued for a long time about the odds. What are the probabilities that there's life elsewhere in the universe lacking any data? There are nothing but arguments and you can argue yourself blue in the face but if you want to answer this question you're going to have to do the experiment. But how can you do the experiment? What is the right experiment? Over interstellar distances there's little sign of Earth. The Sun outshines the planet by more than a billion to one. That's one reason we've not yet seen planets around other stars with even the largest optical telescopes. It's rather like trying to see a firefly perched on the rim of a searchlight. But at radio wavelengths at certain times and certain frequencies Earth's technology outshines all other sources in the solar system by nearly a million to one. The carrier waves of Earth's radio and television broadcasts leak outward in a spherical shell and could be detected by distant civilizations. Radio signals travel at the speed of light, the fastest velocity possible. In contrast, Pioneer-10 has been traveling for nine years to cover some four billion miles but its signal arrives back here on Earth in just five hours. That's one reason NASA researchers think it is communication not physical travel that makes most sense for all species everywhere in the universe. Earth's most sensitive antennas could detect strong signals from halfway across the galaxy. Traveling for just five light hours, Pioneer-10's very weak signal is being used to test NASA's prototype detectors back here on Earth. Barney Oliver, chief of NASA's SETI office. We're looking for a signal now that's coming from the Pioneer spacecraft which is now outside the solar system. It's beyond the orbit of Neptune. Its carrier is a one watt signal and that is about one twentieth of the energy of a candle burning. So we're picking up a really small signal indeed if we succeed. The NASA effort has united Barney Oliver and other longtime SETI proponents with a younger generation of computer builders and programmers including Kent Cullors who despite being blind has worked on the statistics of pattern recognition. The Pioneer-10 is simply a very good example of the kind of signal that we might detect. First of all, although it's kind of close in interstellar terms, it's very weak. One would expect that a civilization would be farther away and stronger. It's near the limit of our detection. It's also a good example of a signal that might be used as a beacon. I'm going to go up a Hertz. Let's see what happens here. The test was successful. Even though the rotation of the Earth and its motion through the galaxy continuously changes the frequency of Pioneer's signal over time. There's a promising candidate right here. I'll have to wait for a few more spectra to see if it... Yes, I think that's it. This bright line that you see going, slanting down the screen is the signal from the Pioneer-10 spacecraft. Now, if you look at other means of communication or making contact, you have to send something. You can't just think about it. You have to send something. And we've investigated all the known particles, and it's certainly true that electromagnetic waves or photons are hands down the best medium to use. So it might seem that we have solved the problem of how to do the SETI experiment. Look for artificial signals at radio frequencies. Unfortunately, it's not that simple. Consider a terrestrial FM radio dial. On Earth, we recognize the channels of our favorite radio stations, defined in megahertz, and can tune from one to the other. Each station has an assigned place on the dial. There, it broadcasts a carrier wave that's always present, no matter the program content. The stations are separated from each other, so that they do not interfere. On the cosmic radio dial, we're also looking for signals artificially concentrated in frequency. But we don't know what frequencies the extraterrestrials may be using. On Earth, between 90 and 92 megahertz, there are just 10 frequencies in use. On the cosmic dial, there are 20 million potentially usable frequencies. In space, the entire electromagnetic spectrum could be used for communication. But in practice, some portions of the cosmic radio dial are harder to use. Lower frequencies that we use for FM and TV broadcasts are overwhelmed by interference from natural sources in our galaxy. So we think it makes sense to listen in the microwave region between 1,000 and 100,000 megahertz. There is a preferred region in the electromagnetic spectrum, we believe, to minimize the energy required to make contact. And that is in the microwave region. And the reason for that is that the noise that would interfere with our transmissions, or theirs, is lowest there. If you go lower in frequency, the noise increases from noises that the galaxy produces. If you go higher in frequency, quantum effects come in and create noise. And in this silent valley between those two walls, communication can be accomplished with the least energy of any region in the spectrum. But even if there are clues about frequency, you still must choose where, when, and exactly how to listen. It's really more like looking for a needle inside a haystack, inside a haystack. For simplicity, SETI researchers compare their experiments using three parameters. In addition to frequency, you have to choose directions to survey the whole sky or to target individual stars. And all experiments will have different sensitivities determined primarily by the size of the radio antenna and by how long they listen. Are there any further clues which can help limit places to look for the needle? Many SETI researchers think there may be, at least in the frequency dimension. Since hydrogen is the most abundant element in the universe, they argue that all intelligent species would think of transmitting their signals close to the frequency at which hydrogen atoms naturally emit radio waves. 1,420 MHz was the first so-called magic frequency to be tried. It was at 1420 MHz that Frank Drake searched in 1960, the first targeted search. He looked at just two stars at this one frequency in a search of moderate sensitivity. In 1973, Ohio State University, with its unusual antenna, began a continuing sky survey. It listens not to single stars, but to the sky passing overhead. It covers 50 channels, but is low in sensitivity. Searches over the last six years by myself and French radio astronomer Francois Bureau at Nancei have looked at 350 stars in the latest targeted search. There has been significant international interest in SETI, with searches mounted in the USSR, Canada, Holland, Germany and Australia, as well as those in the United States and in France. At Nancei, by adapting existing radio astronomical equipment, we've been able to listen to thousands of channels simultaneously. We've achieved improved frequency coverage and a modest gain in sensitivity. As computers advanced, it became possible to construct detectors that were specially adapted to the needs of SETI. One of the first researchers to take advantage of this capability was Paul Horowitz, a professor of physics at Harvard University. The key breakthrough was the ability to scan many channels and frequencies simultaneously. The heart of the system is a so-called multi-channel spectrum analyzer, or MCSA. The one Horowitz uses in his latest system was designed in collaboration with Ivan Linscott and colleagues at Stanford University as part of the R&D for NASA's planned SETI system. Horowitz puts this equipment to practical test in a sky survey known as META, a mega-channel extraterrestrial assay. Drake's experiment could scan only one frequency channel at a time. META can listen to 8.4 million frequency channels simultaneously. It can do in one second what took Drake five years. META, funded by the Planetary Society, is perhaps the ultimate magic frequency machine. But if Horowitz has guessed wrong, he may be looking in the right direction with excellent sensitivity, but just the wrong frequency. No search to date has found any evidence of intelligent life, but SETI researchers feel that their equipment has not been equal to the task. NASA's goal is to extend these previous efforts into a truly comprehensive search. Successful tests of hardware and software in detecting signals like those coming back from Pioneer are but the first steps in a planned 10-year project that will combine the most promising strategies from these earlier searches. The NASA approach that we're developing right now has two different approaches to the search and their complementary. One of them is called the Sky Survey and another is called the Target Search. And the objective of the Sky Survey is to search the entire sky, making no guesses to what are the best directions. To carry out that search in a reasonable length of time, we have to do it quickly, which means we sacrifice sensitivity to signals. We only detect maybe the strongest of the signals that might be there. To cover the sky quickly, the radio telescopes will be driven very rapidly, about 30 times. But it is in frequencies sampled rather than directions covered that NASA's project will dwarf all previous searches. Drake's Lone Frequency is gossamer thin in comparison to the tens of billions of usable frequencies in the naturally quiet microwave window. And even though the Ohio State Survey covers lots of directions and stars, about 60% of the sky overall, its frequency range is also limited. Meta also scans the northern skies, but all of its 8.4 million channels cluster around a single magic frequency, though it will be tuned to others. In contrast, on the frequency dimension, NASA's Sky Survey will cover the whole 1 to 10 GHz window and the entire sky. To achieve this comprehensive coverage of directions and stars, NASA plans to use several telescopes around the world, such as the Spacecraft Tracking System. Each of these telescopes can see two-thirds of the sky, and NASA will move its Sky Survey system between them. The complementary approach is to say, let's concentrate on detecting, with more sensitivity, be able to detect weaker signals. And to do that, we look in a few directions for longer periods of time, and that's called the targeted search. And we pre-select a set of stars that we know a priori to the sun in age and size, and look in those directions with more sensitivity and the ability to detect maybe more complex signals. The targeted search will look repeatedly and at different frequencies at nearly 800 nearby stars. Like NASA's targeted search may start around 1420 MHz, that is 1.42 GHz, but then it will expand to cover all frequencies between 1 and 3 GHz. Like the Sky Survey, the targeted search requires antennas around the world. The giant radio telescope at Arecibo is the largest on the planet, and because of its unique size and sensitivity, NASA hopes to share time to search for the very faintest signals. The large dish can't be moved, so it can only see stars which rotate overhead each day. The antennas at either Ohio State or Nance could be used to observe about 300 stars outside the range of Arecibo. More of the target stars can be seen from the southern hemisphere, so about 200 of them could be observed from the large NASA antenna at Canberra, or perhaps the astronomical facility at Parks in Australia. The last 50 or so stars must be searched from a telescope in the northern latitudes, perhaps rather appropriately from Green Bank, where Frank Drake began the search with Project Osmo. With this unparalleled combination of advanced hardware and dedicated software, NASA's project will have the capability to revolutionize SETI. The targeted search will look for faint signals from what, according to the current understanding of cosmic evolution, seem to be the most interesting sites, and will look at billions of frequencies. The Sky Survey will cover an even wider range of frequencies, yet with sufficient sensitivity to detect a very strong signal from clear across the galaxy. In addition, NASA system will have an unprecedented ability to look for different types of signals that vary over time, not just a continuous carrier wave like that of Pioneer 10 or Earth's radio broadcast. For the first time, we will be able to detect regular pulses that some researchers say will save energy and so be the choice of the transmitting civilization. The interesting thing about the SETI search is that although we claim that we are looking for extraterrestrial intelligence, we can't actually define what an intelligent signal is. And so a starting point is to say we look for a signal which nature will not produce by any mechanism that we understand. Can't colors help NASA design the smart computer algorithms needed to locate weak signals in the many channels of noise? But there will be no humans in the graphics displayed. No human is any more sighted than Kent when it comes to recognizing an extraterrestrial signal. So the eye is not a good detector, first of all, in sensitivity. A thing that you could look at virtually forever and you would never know whether you were seeing random noise or whether you were seeing a real signal. The computer can find signals that are 10 times weaker than that. The amount of data processed per second is the equivalent of the entire Encyclopedia Britannica per second. Of course that's filled with random letters. If that's analogous to noise and what you have to do is find that one pattern that says hi there in all of those random letters that come in every second and then decide whether that pattern could have happened at random because of your statistics of your process or whether in fact that's the real thing. And a human being even if he could perform the task is effectively as a computer couldn't do it for very long. Removing humans from the signal detection process requires very special hardware which is what Ivan Linscott and his colleagues at Stanford University have been designing. And that's like having 10 million ears, each one listening to a particular tone and where all of those are listening simultaneously What we're doing now is taking the technology that exists in this prototype form and casting it in sand. We're building silicon versions of our processors that together with the way we know how to use them will condense both the scale and increase the performance. The process of miniaturization is already underway. 680 of the older large boards will soon be replaced by just 24 smaller boards filled with new chips that are cheaper, smarter and faster and can be more easily transported around the world. No longer will it require you to step in small chunks across the frequency band you can make giant strides and spend lots of time on particular sources or sweep broad swaths of the sky and I think there's where the opportunity lies because it's a prospector's art and you just have to look and take your best bet. In normal operations, the NASA systems will rely very heavily on silicon intelligence with limited human involvement. The computers will have to have sufficient online intelligence to analyze the vast amounts of data arriving every second and weed out the inevitable false alarms before involving humans. For the operational sky survey the sky will automatically be divided into blocks and searched in turn. NASA expects many suspicious signals during each scan but they won't be from extraterrestrials. These events will be compared to the online database of radio frequency interference, RFI, to identify man-made signals. Most of the signals will be explained away until someday, maybe, there might come something more significant. The characteristics of the long hoped-for signal will show some clear sign of intelligence and the computer will have made contact. Then it will call humans to come look and we will know for sure about life elsewhere in space. But before that, increasing RFI poses the greatest obstacle to success. Think of the haystack principle. We're looking for the needle in the haystack. It's a gold needle. Same color as the hay in the haystack. But some joker has thrown so not only do we have to pick up every little straw and sort of crinkle in our fingers to see if it's solid or not but we also have to test every solid one because it might be a bronze one. Since the first SETI experiments we have filled Earth's skies with all manner of satellites and we use them continuously. For this reason, SETI must get started soon or it will become more and more difficult to sort out man-made signals from possible extra-terrestrial intelligence. From the early theories of Philip Morrison and the Osma experiment of Frank Drake, past searches have tantalized as much as resolved. Now, in the proposed NASA search, technology and tasks are beginning to converge. It's millions of times, in fact even billions of times more comprehensive than the sum of all previous searches but falls short by about a million fold from what could be done. So we're kind of nicely positioned in the middle and if we succeed we'll have a cheap way. Searching the Cosmic Haystack with the full NASA system would cost just a few pennies per American per year. Is it worth that price? When I take my paycheck home I certainly pay my house bill and my medical bills and my insurance premiums first and that takes up most of my income but I always save a little bit to buy a book or go to a good movie or go to the art museum or anything that's necessary for us as human beings to keep us human. That's what keeps our curiosity stimulated and our intellect moving and I think we collectively as a human race also need to invest some of our resources into exploration into trying to understand nature as it is around us, nature as it is out there in the cosmos trying to learn who we are and what we're like. I think that's very important to us collectively. I think even if there's a plausible argument for a few we ought to keep looking I'd even go further than that. If there's a plausible argument that there isn't anybody out there bearing in mind that we can be wrong we ought to keep looking because the question is of the most supreme importance. It calibrates our place in the universe it tells us who we are and so it is worthwhile trying to find other civilizations I would say, no matter what. We've hardly searched all the various frequencies the forms of signal the places in the sky from which signals might come and so the fact that we so far have no evidence of extraterrestrial life is not at all discouraging we shouldn't have found it yet we have hardly begun. I think this enterprise can best be understood as a kind of exercise in the archaeology of the future we're well aware of the archaeology of the past we find a site a tumulus or a ruin and we take a spade and we dig into the ground and if you're lucky you discover oer of the Caldees or something marvelous now we never thought that we could examine the same thing in reverse time but in fact in a way we can we know that it's possible that somebody who wants to do it will bring us in of course it is their past but our future even though they're made of different chemistry even though they've never seen our star even though they have nothing biological in common with us they have if they have got radio astronomy if they have the kind of technology we're imagining have very much in common with us the manufacture of a culture the development of a culture which is unmatched among in all the 10 billion species or more that have come to the face of the earth so that's the story and maybe the spade will turn up luckily a good site one day we hope it will it's just a question of being patient when you got the spade and you know the future is there it seems very wrong not to dig