At the recent Astrobiology Science Conference in Chicago, Seth Shostak of the SETI Institute made the intriguing point that intelligent extraterrestrial beings are more likely to be machines—or Artificial General Intelligence (AGI)— than biological organisms.
This perspective might help us answer Enrico Fermi’s challenging question about extraterrestrials: Why haven’t we seen them already? Even if only a small fraction of habitable planets give rise to life, the number of inhabited planets could number billions in our galaxy alone. Assuming a fraction of them develop technology, the number of civilizations should still be considerable. But maybe the span during which technologically advanced life forms remain biological represents only a blip in time.
Shostak explained that machines evolve much faster than humans or other biological life forms; artificial intelligence researchers predict that AGI will overtake the human species within the next 20 to 50 years. To appreciate the fast rise of machines, remember that the first computer was invented in 1945. Whether computers and artificial intelligence can evolve on their own, and perhaps even become conscious, is controversial. Other scientists argue that machines can become only as smart as their designers. However, if we accept Shostak’s argument, the lifetime of a technologically advanced life form based on biology is indeed extremely short, since it will quickly be overtaken by artificial intelligence.
This, in turn, affects our search for intelligent life in the universe. Since most technologically advanced civilizations would be machine-based, it appears we have focused our SETI efforts on the wrong targets. Instead of searching for wet Earth-like planets, we might look elsewhere. AGI, for example, might prefer to settle close to large energy sources such as O-stars and black holes. Figuring out the intent and objectives of the AGI will be more difficult than it would be for biological life forms. Shostak speculated that they might be interested in computer simulations of their ancestors, changing parameters of the universe, or just gathering information from afar.
Would they contact us? And if not, how could we find them? They may not have gotten in touch with us because we haven’t yet passed the threshold to become an Artificial General Intelligence. In that case, the only way to pick up a signal would be to accidentally intercept one that was meant to communicate with an AGI’s “peers.” Such a scenario was proposed at the same meeting by William Bains of the Massachusetts Institute of Technology, as an explanation for the famous Wow! signal detected by the Big Ear radio telescope in 1977. The signal showed the expected hallmarks of a signal originating from outside our solar system, but was received only once.
Still, as Shostak admitted, his was “a very speculative talk.” And so is any interpretation of the Wow! signal.