Seth Shostak and the Odds of SETI Success

"If this project is going to work at all it's going to work before you all become middle-aged." (Seth Shostak)

Seth Shostak thinks that, assuming SETI has the basically right idea, we will likely detect a signal from an alien civilization by 2035. It's important to keep in mind that this probability implies another: it is likely that a signal from an alien civilization will reach us during the next two decades. Indeed, I would argue that Shostak must believe that it is likely that the signal is already hitting us and we just haven't detected it yet. I think this assumption underpins all work in the SETI area: at any given moment it is very likely that a signal, though it may be difficult to detect, is striking the surface of the Earth.

According to the standard model, we're searching about 200 billion stars for somewhere between 10,000 and 1,000,000 "advanced technical civilizations", which Carl Sagan "operationally" defined as "societies capable of radio astronomy". Depending on how many civilizations there in fact are, Shostak argues, and given Moore's law of technological development, we should find the first of them by 2035 at the latest. That is, he is not thinking about the probability that a signal is hitting us; he is thinking about the probability of detecting one of the signals that, he assumes, is hitting us. If there are many, it won't take long. If there are few, it will take longer. But at some point we will have searched the entire probability space of the "cosmic haystack".

But consider our own detectability. Our ambient signal "leakage" from TV and radio (which would be very hard to distinguish from noise in any case) only reaches about 80 light years into space. There are only about 500 sun-like stars in that space and many more stars beyond it. The galaxy is 100,000 light years across. Our signal occupies hardly any of it. It gets even worse when we consider the few attempts at an intentional signal we've sent. In 1974 we sent the "Arecibo Message" for about three minutes. About forty light years away from Earth now, it occupies a tiny sliver of space only three-light minutes long. If it ever does hit a civilized planet, they have to be listening at the exact time that it does. If they blink for three minutes, they'll miss it.

So let's think about the space between us and any one of those "advanced technical civilizations". It is between 4 and 87,000 light years long, with more and more of the stars we're looking at being further and further away. Only .3% of the stars in the galaxy are within 5000 light years of us. And the closer the civilization is to us, the more likely it is that the signal hit us at the time of Socrates and is now a thousand light years away from us and receding! Moreover, we don't know when they might start signalling—a million years ago? A million years from now?

In other words, a million advanced technical civilizations cannot mean a million signals in our sky just waiting to be discovered like any other astronomical object. We can detect the light of stars and galaxies because it has been shining on us for billions of years and will continue to shine for billions more. But any imaginable signal from a technical civilization will have a finite duration. It is, after all, an artifact, a product of the history of the alien civilization. In my view, SETI forgets that alien intelligence, like ours, will be historically situated. When we have not found anything before Shostak's audience reaches middle-age it will prove, as he says, that it's not going to work at all. But we don't have to wait that long to realize that the project doesn't make any sense even on paper.