Technological Singularity

[Matty]

Here's a good explanation of what it is.

So what do you guys think? Will a singularity happen? If so when, and why? If not, why not? What are possible consequences?

[eriador]

Who knows?

But I have a feeling it won't happen while I'm alive, so from my point of view, it will never happen.

[Paul]

ooh, do i smell another vernor vinge fan?

oh and eriador, i think it will DEF happen within our lifetime. Well, unless global warming kills off most of the population, causing our rate of progress to slow down...

Just look at how fast we are progressing.

-The internet of 5 years ago can easily fit multiple times on just one of googles new data centers (they have sooo many). Ive also heard that the number of internet sites doubles every year.

-Computers are getting faster and faster, Moore's law (in some form anyway) shows no signs of slowing.

-while some people have been throwing around the term "web 3.0" (idiots), really the whole web 2.0 is still in beta.

-people are already starting to confuse virtural worlds with the real one. Just look at all the people that are hooked on Wow.

Big things are brewing, i have no idea what exactly what they will mean, but they will be big. Mark my words, assuming, there isnt a MAJOR population drop and/or major economic depression, i think the human race will reach some sort of singularity withing 30 years.

[AnthonyByakko]

I'm with sparky.

I've always planned on, before I die, interfaces being created to electronically transfer the content of an organic neural net to a digital format; essentially, putting one's brain on hard drive. I plan on this occuring sometime by 2085; which will make me 100 years old, but this number seems larger in our current time. I suspect that, with the care I take of myself and advancing medical/scientific technology, I can reach that age and thus that "singularity."

I have other plans, of course, for after this happens; depending on where the technology goes from it. Do I put my brain from the hard drive into a newly cloned/fabricated human body every time one wears out? Do I exist on the "hard drive" in a synchronous virtual word? Is some non-human form created to "host" a brain and provide non-virtual, physical interaction with the real world? Who knows. All that matters is that it means I'm never going to die. (But of course you have to have contingency plans for other outcomes. Should for some reason technology not live up to my projected growth, I suppose I can be reincarnated and wait until it catches up.)

Then we could, much more easily, populate other worlds as Hawking recently suggested (in a bit of a "Captain Obvious" moment.)

[Dr. Mobius]

In futures studies, a technological singularity (often the Singularity) is a predicted future event believed to precede immense technological progress in an unprecedentedly brief time.

I didn't get much past the first few lines, but in my opinion didn't we already have the singularity with the advent of computers and microelectronics? Or is this the next one we're looking for?

[KennEnder]

Somehow I have the impression that arriving at The Singularity implies that there is no time delay between the "need" for a technology and its fruition... so we haven't quite made it there yet. (although the time delay between "significant events" is certainly on the order of 10^1 vs. 10^10 years now!)

[eriador]

ooh, do i smell another vernor vinge fan?

oh and eriador, i think it will DEF happen within our lifetime. Well, unless global warming kills off most of the population, causing our rate of progress to slow down...

Just look at how fast we are progressing.

-The internet of 5 years ago can easily fit multiple times on just one of googles new data centers (they have sooo many). Ive also heard that the number of internet sites doubles every year.

-Computers are getting faster and faster, Moore's law (in some form anyway) shows no signs of slowing.

-while some people have been throwing around the term "web 3.0" (idiots), really the whole web 2.0 is still in beta.

-people are already starting to confuse virtural worlds with the real one. Just look at all the people that are hooked on Wow.

Big things are brewing, i have no idea what exactly what they will mean, but they will be big. Mark my words, assuming, there isnt a MAJOR population drop and/or major economic depression, i think the human race will reach some sort of singularity withing 30 years.

vernor vinge??

Perhaps my resistance to accept it coming in my life reflects my fear of singularity. However, I think that instead of the rate of advance increasing exponentially (fig. A) I think that the growth will asymptotically approach the theoretical maximum of growth asymptotically, tracing a sigmoid graph (fig. B). In that case, we would currently be in the steepest part, but both ahead of us and behind us, the rate was and will be slower. That's because I think there is a theoretical maximum to how quickly we can develop technology, and the sigmoid curve fits the data we currently have and fits with my thoughts on maximum rate.


Fig A:


Fig. B:

[fawkes]

I agree on some level with Eriador. At some point, if our technological advances are not growing exponentialy, there will be a plateu, and there will be nothing else to invent/improve. However, I find it hard to believe that it will happen in our lifetime. There are still far too many scientific discoveries to be made for there to be a technological singularity anytime within the next hundred years or so.

[Oliver Dale]

I certainly think there are arguments to be made on either side. But, even if you accept a sigmoidal approximation of technological advancement, there is no guarantee we're on the steepest portion of that right now.

Some of the problem comes from trying to model this growth. How do you define technological advances quantitatively? Computer processing power? At a certain point you're guessing, and that becomes dangerous when you deal with exponentials.

Instead, you have to treat this philosophically. And the idea is hard to deny. There very well may come a point in time at which computers develop a synthetic intelligence capable of self-improvement. When that happens, it is hard to imaging anything becoming a barrier to advancement, unless you're talking about, you know, knowing everything. Or, global destruction.

[eriador]

I certainly think there are arguments to be made on either side. But, even if you accept a sigmoidal approximation of technological advancement, there is no guarantee we're on the steepest portion of that right now.

We're definately on the lower half, where it is concave up, as all the current data points that way.

Some of the problem comes from trying to model this growth. How do you define technological advances quantitatively? Computer processing power? At a certain point you're guessing, and that becomes dangerous when you deal with exponentials.

Instead, you have to treat this philosophically. And the idea is hard to deny.

I just think that there is a theoretical maximum of how fast we can discover technology. Hence the asymptotic aproach represented by the sigmoid.

There very well may come a point in time at which computers develop a synthetic intelligence capable of self-improvement. When that happens, it is hard to imaging anything becoming a barrier to advancement, unless you're talking about, you know, knowing everything. Or, global destruction.

Umm, haven't humans already gotten an intelligence capable of technological improvement? How would a computer intelligence that could invent things be any different from human invention? You're just applying the reasoning justifying humans reaching singularity, but replacing humans with computers. My comments about a theoretical maximum still stand, unless I misunderstood what you're saying.

[zeroguy]

Some of the problem comes from trying to model this growth. How do you define technological advances quantitatively? Computer processing power? At a certain point you're guessing, and that becomes dangerous when you deal with exponentials.

Instead, you have to treat this philosophically. And the idea is hard to deny.

I just think that there is a theoretical maximum of how fast we can discover technology. Hence the asymptotic aproach represented by the sigmoid.

I don't know if he was contesting that, but it's just that the "data points" are so arbitrarily chosen it makes fitting a curve to the data meaningless, since the data is so fuzzy.

[eriador]

You win on that point, but I was working on the assumption that most people seemed to be working on, which was that technology has been increasing exponentially.

[AnthonyByakko]

Perhaps you didn't hear me.

ROBOT BODIES!

[hive_king]

Wow. I first read that as

ROBOT BOOBIES!

[AnthonyByakko]

Wow. I first read that as

ROBOT BOOBIES!

Technology is sweet.

[eriador]

ROBOT BOOBIES!

That will be the end of all hope for the continuation of our species! But I must say that I am all too eager to see "robot boobies."

[Matty]

Umm, haven't humans already gotten an intelligence capable of technological improvement? How would a computer intelligence that could invent things be any different from human invention? You're just applying the reasoning justifying humans reaching singularity, but replacing humans with computers. My comments about a theoretical maximum still stand, unless I misunderstood what you're saying.

The assumption on which the singularity is based is that computers will one day be faster and/or better at inventing than humans. When that happens, they will be able to improve their own intelligence more rapidly and more effectively, which will lead to intelligence that is even better at modifying itself, and... pop goes the weasel.

[luminousnerd]

I read the Wikipedia article and it doesn't make any sense to me. Is it possible to describe to a simpleton like myself?

[jotabe]

The point is that the more we know, the more we can know. Also, the more we know, the more means we have to know even more. Hence, the defenders of this theory say that the growth of total human knowledge is exponential.

When you have an exponential function, you can roughly approximate it in two parts: a low growth one, and after a "take-off" point, an almost vertical one. This take off point, for the assumed exponential growth of knowledge, would be the technological singularity.
They say that this technological singularity would happen when we finally manage to create AI, an AI capable to improve itself.
We know that machines can do what we program them to do, but a lot faster than we could. If they could program themselves, debug their own code, improve their own hardware, making themselves more efficient... that would be cool.

Personally, i believe all of this is nonsense. Wishful thinking. It's true that the achievement of AI will be great, and it will provide a huge improvement in the speed of our technological development... but that doesn't mean that is unlimited. Sooner or later, AI will find techonological problems that will be beyond their means. Also, i don't think that a singularity will ever happen... true exponential growth behaviours have never been observed in Nature (even for the singularities inside Black Holes, nature provides a "cosmic censorship" so we can't see them: the event horizon). Sooner or later the apparently exponential growth stagnates.

[Boothby]

So, after you have transferred your intelligence to a hard drive sometime in 2085...what happens when someone writes a virus that infects the hard drive, turning all the memories involving bicycles into memories where you continuously smack yourself in the head with a trout?

And then the spindle bearings fail, and you forgot to make a back-up... AGAIN!?!

[Guest]

So, after you have transferred your intelligence to a hard drive sometime in 2085...what happens when someone writes a virus that infects the hard drive, turning all the memories involving bicycles into memories where you continuously smack yourself in the head with a trout?

And then the spindle bearings fail, and you forgot to make a back-up... AGAIN!?!

Like when your video game save file gets corrupted?! OH NOEZ!!!!!!!!