Saturday, July 14, 2007

Tipler's good news & bad news about interstellar travel

I've just stumbled across an old paper in The New Scientist written by a scientist named Frank Tipler in the early '80s that offers an interesting perspective on interstellar travel, and I've been thinking about how to use it in a story, because it remains fairly minimally exploited.

If you accept Tipler's view of the future, there is good news and bad news regarding outer space. The good news is that interstellar travel and colonization from Earth is not only possible, but likely, sometime in the next few centuries.

The bad news, however, is 1) it won't be people doing the colonization, and 2) we won't be meeting any aliens along the way.

Tipler starts from the assumption that fairly soon (within a few hundred years, at least), we should be able to create self-replicating, sentient "von Neumman machines" that are able to travel to other star systems (on their own time, of course, not limited by human life spans) and begin reproducing themselves there.

Based on the assumption that any intelligent species would be able to do the same if they got to the same point in the development of consciousness, he points out that if there were a significant number of other consciousnesses in the galaxy, we'd have already seen their VN machines tearing the place apart by now. Therefore there must not be any.

He has a second reason for doubting that there are any other sentiences out there; he points out that the famous Drake formula that assumes other sentiences are common (# of intelligent species = number of stars that form * percent that have habitable planets * percent of habitable planets that evolve life * percent of planets with life that develop to sentience) may be making a big leap when it assumes that the likelihood that sentience will be the natural result of any evolutionary process. He argues that in fact the likelihood for the evolution of sentience is in fact infinitesimal.

I'd say that his assumption of the likelihood of "von Neumann" colonizers is his big potential weakness. It does seem likely based on the way that info- and nano-technology appear to be evolving. But until we're actually sending von Neumann colonies to Jupiter, it's far from a slam dunk.

But for now, let's pretend he's right. The idea that the colonizers of outer space will be machine intelligences rather than people is not so discomforting to me as it is to other people. After all, I'm quite certain that neither myself, my son, or his grandchild(ren) if he has any will actually be traveling to Alpha Centauri and beyond. So what do I care if it's my great-great-great-granddaughter or a machine she makes that colonizes the distant stars? (I know, I know, what do I care if anyone does or not? But I do, okay, so shut up.) The idea that there wouldn't be any other intelligences out there seems a bit lonely, but once your von Neumann machines started evolving, competing, spreading and, speciating the galaxy would likely get to be a pretty lively place after a few million years or so, and by then perhaps no one would even know where it all started.

But of course I'm looking at this from the point of view of a scifi writer. There's one major challenge here, which is that, as I said before, fiction tends to be about people. I have yet to be aware of a major work of science fiction that has no characters at all that are either human, or at least very human-like. That's not surprising given my assumption that science fiction is about people. When you're describing a sentience different enough from us, readers will just no longer be able to identify. But it is almost certainly true that Our New van Neummanist Overlords will be almost nothing like humans, sentient though they might be. It is exactly their non-human qualities that will allow them to go to outer space. And I'm not just talking about the physical part, but psychological differences, for example the immunity to boredom that would be necessary to travel for what might be millenia or more.

Of course, there is the possiblity that interstellar colonization will occur in two stages, with the machines going first, then setting things up for the humans much later. After all, interstellar travel is far more compelling if we know there is someplace to go. But then, if the machines are smart enough to colonize outer space, it's open to question exactly what they'll need us for. By the time any humans got to interstellar habitations that von Nuemann machines set up for us, we'd likely find that at best we'd be thought of as adorable housepets. At worst we'd be a nuisance to exterminate. (Yes, the dear, late Saberhagen addressed this in the Beserker series.)

So how do we tell the story of a future in which humanity was just an imperfect step way back down the complexity ladder? I don't know yet, but the possiblitities are too tempting to ignore. When sentiences are able to live hundreds of thousands or millions of years, and can be confined to centimeter-long vacuum-and cosmic ray-proof packages, then many of the limitations that make the stars so distant become trivial matters.

Just as a suggestion, what about a "starship" the size of a chiclet composed of hundreds or thousands of sentient intelligences within its own memory guiding its travel at, say, a few psol or so. A bizarre paper by a man named Jim Walker suggests that for an artificial intelligence time could be slowed down by 'ceasing to exist' for centuries or more at a time. Imagine, for example, experiencing only one second or so out of every few hundred years. If an intelligence was experiening that, it could travel at what appeared to be many times light speed.

It would be a bit of a stretch, but imaginable that the future von Nuemann consciousnesses could compose themselves out of a large number of very human-like intelligences, that would experience an artificial reality within their virtual environment, potentially even fighting, eating, drinking, screwing and reproducing within their environment. Really, if you think about it, not that much crazier than some of the stuff that Charles Stross comes up with. And I haven't thought about it long, other possibilities must occur.

No comments: