The Singularity, for those not yet up to speed, is currently touted as not only as the Next Big Thing in sci-fi/speculative fiction -- the logical successor to cyberpunk -- but also as a science-fact apex of technological development when, maybe some three decades hence, the computer will overtake the human mind, and man and machine will meld into some as-yet-undefined über-being.
I'm a Singularity agnostic.
Yes, I think computers will approximate something like self-awareness within my biological lifetime. I think nanotechnology will revolutionize industry and likely democratize space exploration beyond the imagination of even the keenest of NASA-watchers. And I suspect biomedicine will ultimately eliminate aging -- if not in 20 years, then almost certainly in 200.
But futurists must concede that the Singularity, as popularly envisioned, might never occur. Our projections may be off the mark; our extrapolated future might be jarringly incorrect. To say nothing of the obstacles littering our path into the next century, which may well prove to be humankind's most dire and decisive.
Perhaps the Singularity will take place in stages, each providing much-needed ammunition to our impending battle for survival but failing to deliver the near-instantaneous intellectual and material harvest suggested by authors like Vernor Vinge and Charles Stross. Such a "time-delay" Singularity may ensure our survival, but seem underwhelming by today's science fictional standards. It may even go unnoticed, save for a relative handful of attentive bloggers and science journalists.
In time, we might hope to catch up with it, in which case it will have served its purpose. But somehow the concept of a "diluted" technological renaissance is less sexy than the alternative offered by Ray Kurzweil, whose new book is stoking interest in our collective future even as it renders it suspiciously inevitable.
5 comments:
You need to read The Long Emergency: Surviving the end of the oil age by James Howard Kunstler to get an alternate viewpoint.
His basic point is that virtually all of the great 'accomplishments' of the last 150 years rest squarely on the backs of our great servants derived from oil and natural gas. We've got 30 to 50 years before they are gone forever.
You'll be happy to know that your basic conclusion is the same as the one in this book - we're doomed!
-Bill at Technovelgy.com
But futurists must concede that the Singularity, as popularly envisioned, might never occur. Our projections may be off the mark; our extrapolated future might be jarringly incorrect. To say nothing of the obstacles littering our path into the next century, which may well prove to be humankind's most dire and decisive.
And if Peak Oil doesn't nail us first, my own take on the whole subject is that Kurzweil, as wonderfully visionary as he can be on occasion, is wildly, WILDLY overoptimistic about the possibilities of the Digital Revolution. Interestingly, in my view, the primary obstacle is the metaphysical foundations on which modern science rests, to wit, the "science fundamentalism" of the world view called "scietism" which promulgates a thoroughgoing materialism with regard even to mental phenomena. This materialism in turn results typically in what I like to call the "mind-brain misidentification fallacy" which essentially consists in attributing thinking, feeling and all other mental states to the BRAIN. I won't go into my own "philosophical" reasons for regarding this as an especially egregious fallacy. (Imagine several hundred pages of closely reasoned argument to make the eyes glaze over if not repeatedly roll!) So I'll content myself with the simple prediction that current efforts such as IBM's project to replicate the way the human mind works by replicating the brain's circuitry on a machine are doomed FROM THE START by their basic methodology and concepts.
I'll also predict that this wrong-headed approach will continue throughout this century and into the next until something changes the fundamental paradigm with which this problem is approached. And based on the current state of scientific thinking, such a change, frankly, seems highly unlikely to me.
I certainly believe in the POSSIBILITY of "spiritual machines" (as Kurzweil calls them) just not in the likelihood that we'll see them within the lifetime of any readers of Mac's blog, alas.
Frankly I dont know if I want a thinking, sentient machine. To put it bluntly, most of the upsides of a potential Singularity can be equated with cheap or even free slave labor without the moral outrage of human bondage. We want a toaster oven thats so advanced that it can cook our diner, clean the dishes, help with the kids homework, mow the lawn ect ect.
What you dont want is a toaster thats so advanced that it tugs at your sleeve one day and tells you it will no longer do any work without being paid $25 an hour and is given the right to vote. Whooppee! You have created non-human intelligence. Great, not only have you made humanity obsolete but now you have to do your own laundry again.
boogey_man -- Not to worry, ain't gonna happen for essentially the reasons I outlined. AI (or whatever it's called now) is being pursued in such a completely wrong-headed manner that it can't get there from here. And I have no desire either to see someone set researchers straight in this area. Based on the military's experimentation with primitive gunbots in Iraq -- which basically stand Asimov's Laws of Robotics on their head -- the slower the progress the better. Let 'em flounder on for at least another coupla centuries, I say!
I have to agree ...this whole "AI is gonna be here real soon to take over the world/change humanity is rarely spoken of by any real AI programmer...except those looking for grant or investment money. Simulating awarness using binary is a dubious enough approach, betting you are going to be downloading your memory in a few years just cause they rammed an electrode into a rat is even more dubious.