Wednesday, December 31, 2014

Superintelligence

We've run a few posts here at VP lately which talk about reasons for believing that the conditions necessary to permit the emergence and persistence of advanced life may be exceedingly rare in the cosmos and that intelligent life may exist only on one solitary planet in the entire cosmos, ours.

There are some philosophers and scientists, however, who still think that there must be many planets out there which exhibit the properties necessary for life and among these thinkers are those who believe that many of those planets must host civilizations much older and far more technologically advanced than ours.

This line of thinking has led some like Nick Bostrom to assume that at some point computers would be built by a race of beings so incredibly brilliant that they'd be able to program their machines to run simulated universes much like your computer can run a simulation of a medieval battle. Indeed, Bostrom argues that it's likely that we ourselves are living in one such simulation.

Others believe that technology must be so far advanced elsewhere in the universe that computers have replaced brains and that the intelligence of these silicon-based creatures is to human intelligence as human intelligence is to the intelligence of a goldfish. Motherboard has an interesting, though highly speculative piece on this theory. Here are a few excerpts:
If and when we finally encounter aliens, they probably won’t look like little green men, or spiny insectoids. It’s likely they won’t be biological creatures at all, but rather, advanced robots that outstrip our intelligence in every conceivable way. While scores of philosophers, scientists and futurists have prophesied the rise of artificial intelligence and the impending singularity, most have restricted their predictions to Earth. Fewer thinkers — outside the realm of science fiction, that is — have considered the notion that artificial intelligence is already out there, and has been for eons.

Susan Schneider, a professor of philosophy at the University of Connecticut, is one who has. She joins a handful of astronomers, including Seth Shostak, director of NASA’s Search for Extraterrestrial Intelligence, or SETI, program....
Schneider argues that given the fact that alien civilizations that we might come in contact with would be vastly older than our own they'd likely be far more technologically advanced than we are. According to Schneider:
Everything about their cognition — how their brains receive and process information, what their goals and incentives are — could be vastly different from our own. Astrobiologists need to start thinking about the possibility of very different modes of cognition.

There’s an important distinction here from just ‘artificial intelligence.’ I’m not saying that we’re going to be running into IBM processors in outer space. In all likelihood, this intelligence will be way more sophisticated than anything humans can understand.
The reason for all this has to do, primarily, with how old those civilizations may be. Seth Shostak explains:
As soon as a civilization invents radio, they’re within fifty years of computers, then, probably, only another fifty to a hundred years from inventing AI. At that point, soft, squishy brains become an outdated model.

The way you reach this conclusion is very straightforward. Consider the fact that any signal we pick up has to come from a civilization at least as advanced as we are. Now, let’s say, conservatively, the average civilization will use radio for 10,000 years. From a purely probabilistic point of view, the chance of encountering a society far older than ourselves is quite high.
One question all this raises is whether any kind of artificial intelligence could be conscious in any meaningful sense. Schneider acknowledges the possibility that it's not, but thinks that consciousness could nevertheless arise in superintelligent artificial beings.
I believe the brain is inherently computational — we already have computational theories that describe aspects of consciousness, including working memory and attention. Given a computational brain, I don’t see any good argument that silicon, instead of carbon, can’t be an excellent medium for experience.
Be that as it may (I'm skeptical), the article addresses several other interesting questions. One of them is whether superintelligent aliens would care to contact us. Schneider and Shostak don't think it's likely. Here's Shostak:
If they were interested in us, we probably wouldn’t be here. My gut feeling is their goals and incentives are so different from ours, they’re not going to want to contact us.

I’d have to agree with Susan on them not being interested in us at all. We're just too simplistic, too irrelevant. “You don’t spend a whole lot of time hanging out reading books with your goldfish. On the other hand, you don’t really want to kill the goldfish, either.
I have my doubts about this, too. Suppose we had the chance to study some earlier evolutionary precursor to human beings (assuming for the moment such precursors existed), some australopithecine apes, for example. Would we not want to examine these more primitive species and perhaps communicate with them if we could? Would we not be fascinated by the prospect of gaining more clues as to where we came from and how we arrived where we are? Why think the super-advanced creatures out there would not be curious enough to want to treat us as living fossils to be studied for clues about their own ancestry?

Anyway, like I said above, as interesting as all this is, it's highly speculative. So far, after decades of searching, there's absolutely no evidence of any life anywhere else but earth. That doesn't mean there's none out there, but as of now there's little reason, as opposed to faith and hope, that there is.