So this is my first post here, basically just testing the waters, so to speak.
A year and a half ago I read Kurzweil’s “Singularity is Near” and pretty much bought into all of his claims hook line and sinker. I still agree with some of the ideas he puts forward (Transhumanism and life extension being desirable mainly), but looking back at his predictions, they get increasingly outlandish the further out he predicts (like having foglets by 2045).
I’m just wondering what your thoughts are in terms of the plausibility/realism of a singularity-type event involving ‘recursive self-engineering’; does anyone here see that as remotely plausible?
Remotely possible? Of course. But, the very definition of a Singularity includes the concept that it is the point where our conceptual models break down, and therefore we do not have the tools to make predictions “past” the singularity.
Will there be a point at which technology takes off in directions we cannot foresee at this time? Maybe. I don’t think anyone could have predicted today’s world at the time of the Steam Engine Singularity.
Is the “Nerd Rapture” possible? I sincerely doubt it.
I haven’t read the book in question, but am familiar with the concept of a technology singularity. Is it total bull? No, I don’t think so. The basic premise of an increased and ever increasing rate of technological change and convergence of multiple technologies seems pretty solid, since we are seeing that happening. And the rate of change just in my own lifetime has been pretty staggering, and continues to increase from my own perspective, at least in several vertical areas. Will this lead to the proposed singularity? No idea, but probably not. But even if we don’t get all the pie in the sky of The Singularity™, some pretty amazing stuff is coming down the pike in the next 20-30 years, and should be quite a ride, regardless.
Pretty much. Only this time the shiney new thing is computers. (I actually have a half-baked theory that the only significant technological changes are those which improve the production and distribution of food, since even technomage geeks need to eat.)
Ok, this is pretty different from my understanding of the word ‘singularity’. From what I can tell, ‘singularity’ doesn’t just refer to any disruptive technological advancements. People like Vernor Vinge and Kurzweil use it to refer to a fairly specific theory of events: The core being that once ‘Strong’ AI (here meaning AI on the level of humans or above) is built, that same AI should be able to re-engineer itself to be smarter, which allows it to become even smarter, on and on at an intensifying pace. This is the fundamental claim that Kurzweil uses to support his predictions that AI’s will solve intractable problems like aging and long distance space travel in a fairly short amount of time.
It is much easier to imagine strong AI than to actually build it. I took my first AI class at MIT in 1971. One of the books we used was from 1959. We heard about a whole bunch of problems, almost all of which have been solved now, including generating routes for travel, chess, solving calculus equations. I’d say we are not much further toward true AI than we were 40 years ago. Bigger computers ain’t going to hack it. When the 386 came out the USA Today said that AI was on the brink of being solved now that we had 16 bits.
Computers design themselves already. Much of the work that people used to do, like laying out gates, is done automatically. But it still takes architects and top level designers. Computers help.
What we are going to see is more computer help, kind of like a Jeeves to our Wooster. Computers will make reservations for us, remind us of the date, suggest getting flowers, but it won’t help us make small talk.
I don’t think any singularity we’re going through this century is all that much different from the singularity we’ve already gone through, and we’ve mostly handled that one pretty well.
Food, or energy. I recall in James Howard Kunstler’s Too Much Magic, he was at a conference with some computer scientists speculating on the idea that a strong AI might want to do a Skynet and eliminate/replace the human race. Kunstler asked, “But, then, where would its electric power come from?” Stumped 'em – of course there are conceivable answers, but the point is that the question had never even occurred to these experts before; they had always thought of electricity as something that just comes out of the wall.
That’s the best put-down I’ve ever seen! And it shows why all the talk of miracles is such bunk.
Yeah, computers are getting better.But they are nowhere near the level of true intelligence.
The techno-geeks love to masturbate to stories of how computers have improved at playing chess, while comparing the size of their [del]dicks[/del] processors. .
But they forget that somebody still has to plug the computer into the wall, which means that somebody had to approve the architectural blueprints, and somebody else had to dig the ditch from the building to the street to lay the electric cable , which means that he had to sweat, and after all that sweating he took a break to eat and drink, and then after eating he had to shit, which required somebody else who knows how to build a toilet and lay the sewer pipe to the waste treatment plant, and somebody else who passes a local ordinace requiring the use of low-flush toilets only, and somebody else who knows how to advertize toilet paper and air fresheners.
There’s an old saying: Both a computer and a person can win a great game of chess.But only a person can enjoy it.
Life ain’t gonna change much in the next 50 years. We’ll have self-driving cars, we’ll watch 3d holograms instead of TV, and we’ll have more robots. But you’ll still have to stand in line at the airport and pay $20 to check your bag, even if it is a robot who stamps your biometric passport and puts your bag on the conveyor belt.
Porn may get more better, too…but when you want to really want to make love, you’re gonna have to have learned people skills, not computer skills. Just like back in the primitive days of the 20th century when computers had green screens and ran on DOS…
Personally I don’t believe that AI at a level of humans are above is going to appear, at least not within our lifetimes. As Voyager mentioned, there have been massive increases in computing power over the past few decades, but they have not been matched by huge progress towards better AI. Jaron Lanier’s book You Are Not a Gadget delves into the issue at some length.
[QUOTE=chappachula]
Life ain’t gonna change much in the next 50 years.
[/QUOTE]
No more than, oh, say the last 50 years. Trust me, when you actually sit down and think about it, the change is pretty profound, and I see nothing to indicate it’s going to slow down…quite the opposite, really.
And yet, almost unnoticed seemingly, they now have a computer that can beat the best human player at Jeopardy, which is a huge achievement. I think some of you don’t realize how far we’ve come in a fairly short time, or how many different technologies are converging.
Watson has vast databases that it can access quickly and guess at the most likely answer. It can ring in mechanically fast. So, yes, it is better than most humans at Jeopardy, but I find my car GPS device to be far more useful and amazing.
Right, but Kurzweil’s argument is that we are going to get a more sophisticated understanding of the human brain, and through that understanding be able to engineer AI’s based on our neurological architecture. This is his justification for saying that AI research will take such a big leap in the near future.
Now, the issue with this is that there is no certainty that we will have a full understanding of the workings of the human brain in the next 50 years, let alone the next 20. However, if you allow for the possibility of a full working model of the human brain, Strong AI becomes a much more realistic possibility.
The biggest changes were between 1900 and 1950: electricity, flush toilets, refrigeraters, and medical care. Deep, profound changes. Life went from chaos to stability. (in 1899 every childbirth meant possible death for the mother, and many children died of diseases, food supply was erratic and dependant on the weather)
Between 1950 and 2000 we got television and the internet. Also deep changes, but not so profound. TV brought us closer together, so racism became unacceptable. Internet makes life much easier, etc, but not as profoundly easy as the refrigerator.
Between 2000 and 2050–I expect life to get easier, more connected, more robots, etc…But not as profoundly different as the changes my grandparent lived through.
Trust me, when you actually sit down and think about it, the change is pretty profound, and I see nothing to indicate it’s going to slow down…quite the opposite, really.