The Singularity is near

Ray Kurzweil Interview

The link above is a link to an interview of Ray Kurzweil on NPR. The interview is in audio (mp3?) format and it's about 35 minutes long, but it's definitely worth checking out if you find the time.

In the past few years I've been doing some reading about nanotechnology, biotechnology and about how technology is levelling our world. Based on things I've been reading, it's hard to not conclude that we are entering a new era.

Some of the things Ray predicts are incredible, but he explains his predictions well. If you're unfamiliar with his name, he's an MIT graduate who is now an inventor and author. He's written several books and he's contributed a lot to our world in terms of technology. He invented the flatbed scanner we are all familiar with, he's worked a lot with computer AI and text-to-speech pattern recognition programs, worked on the first commercially-marketed large vocabulary speech-recognition, he's started 9 companies, created the first multi-instrument musical synthesizer -- he's done a lot for one guy. Make no mistake, he's definitely a genius.

What do you think about his predictions?

 

Log in or Sign up to hide all adverts.

That they are somewhat overblown.
That the singularity, whilst a nerds wet dream, is unlikely to happen for 20 or 30 years at least.
That Drexlerian nanotech is vastly oversold.
That, given the still current lack of an actual AI entity, there is a lot of work still to be done.

 

Log in or Sign up to hide all adverts.

Right now on C-Span 2, Ray Kurzweil is doing a presentation on his book, The Singularity is Near, if anyone wants to catch some of it. Pretty darned interesting.

- N

 

Log in or Sign up to hide all adverts.

Here's a link to a self-interview with Ray Kurzweil.

If read completely through, he makes some pretty powerful and undeniable points, especially in reference to the exponential quality of the rate of evolution.

It seems that from the very beginning, biological evolution has been exponential. Very slow. And then comes along DNA, which sped up evolution by bounds and lleaps, and then eurkaryotes and multicellular organisms and so on. Builds on top of itself.

And now evolution in technology is so rapid, we can actually see it change right before our very eyes. And no, the speed of change is not remaining the same. It's getting quicker and quicker and at a quicker and quicker rate. Pretty scary stuff if you ask me. And the more I read the news on the Internet about new processors and applications and AI, the more I believe that not only is it going to happen, but it's probably going to happen sooner than predicted.

 

Ray's a bright guy... I met him at a NAMM show in 1990. He impressed me with his ability to describe mass parallel branching and adaptive synthesis in AI before we had even moved to 32 bit PC's.

Since then, volumes have been written about self-aware adaptive machines, but none can evolve yet without Human input.

 

I'm just suscribing (writing this) so I can find this tomorrow to listen to and read the interviews. Anyone care to quickly summarize the definition of singularity that you are referring to?

 

The Singularity refers to a time in the future where the evolution of technology is so fast that we cannot predict or even understand with our brains what the future will look like. It got that name from the singularities of black holes. Beyond the singularity of a black hole, you can't see anything. Or something. I really don't know shit about black holes, but that's the gist.

 

Thanks. It's exciting, isn't it?

 

Shit, I want some of what Kurzweil is on. He should sell it to people, best drug on the planet.

A sceptical look at what Kurzweil is saying in the self interview:

For starters, he seems to confuse nanotechnologies. There are the manufacturing kind, which are like miniature factories (call them assemblers), and then theres the kind of stuff like carbon black, where the material does what it does because its nanometre size anyway. He refers to nanotechnology based solar panels, whatever they are. true, we are currently looking at making efficient solar panels by using nanoscale bits of material, but there is no guarantee that it will work, nor is that the same as making assmeblers.

If we manage to get past silicon. The technologies to do so are barely in the laboratory stage, and again, there is no guarantee of their succeeding.

Call me a bit cynical, but I dont see them as intelligent, I seem them as well trained. There is a difference.
Now, self evolving programs are in development, but I have extreme trouble seeing that they will get to what he is looking for.

In case you havnt noticed, my main point is that although much of what he is saying is sensible, I disagree both with the timetable and with whether or not much of the technology is practicable. Take assemblers for example. We do not yet have more than a few proposals as to how to make them. The prototypes for their manipulative capabilities are still in the laboratory, and as far as I am aware many questions still need to be answered about them.
So, with regards to nanotech, despite the billions of dollars that have gone into it in the past 10 years, I see no reason why we will have assemblers and nanotech inside our bodies in the next 20 years. 40 years, I think there is a good possibility.

What gets my goat is simply that he writes "We will", and that he is putting himself forwards as the great prophet. Which he has been for nanotech etc, but given that the area is 10 years old at least, indeed, Drexler published his first work on it over 20 years ago, and SF has been exploring it continuously for 10 years and more, I think he is being a bit hubristic.

Finally, he seems to completely ignore the politics and economics of the possible singularity. Which in one sense is fair enough; if it is that much of a singularity, we wont have a clue about what happens afterwards. Those of you who actually read anything will know about the Vernor vinge book where the singularity has actully occured. "Marooned in real time", where the characters have been trapped in no-time bubbles, kind of like suspended animation, except its time that is frozen, not them. So that the bubbles pop, and they ocme out and find that they are in a future where everyone has disappeared. The singularity, with incredibly advanced techynology has happened, leaving them behind, and they dont know where everyone has gone.

What things will be like after a singularity will matter in part on the realtive strengths of the entities involved. Access to matter and energy will still be important. If intelligence is really that important, actual warfare may well be limited, but conflict may still be likely to occur.

 

I'm not sure what you mean by this...

Do you really doubt that we'll get past silicon?

I see your point. Obviously it's a semantic argument, but the kind of intelligence I believe he's talking about in his context fits into the description of "weak AI". AI is still in its infancy, but strides are being made.
IBM - Advanced Voice Recognition

I know. Itseems quite incredible to me too. But then again, he's a lot smarter than I am. And he knows a lot more.

Those assemblers thought up by Eric Drexler do not look like they would work. It seems he's just trying to take an idea like a robotic arm and shrink it down to the nanoscale and make it work, which doesn't make a lot of sense because the world behaves much differently at different scales. Your new obstacles are governed by different forces, like nuclear and the pull of different molecules and atoms, while gravity won't be so much of a concern.

In reality, however, real working nanobots likely won't look like anything man-made. They will probably look more like imitations and modifications of the machinery that already exists in our bodies (enzymes, DNA, organelles), except of course, they will do what we want them to do.

Never heard of it.

 

Singularity is the state where all divisions, whether abstract or not, are condensed into the same point and become one (ie. singular).

 

That's one definition of Singularity. However in the context of this thread, it means, well, this is how Wikipedia puts it:

It's a relatively loose term though.

 

The idea is very cool. As to whether it'll happen - for myself, I'd rather wait and see. Doesn't a lot of it depend on undiscovered physics working conveniently when we do discover it?

Most humans need human input too

Please Register or Log in to view the hidden image!

 

Well, carbon black is essentially tiny bits of carbon atoms stuck together in a structure a bit like graphite, but more fragmented, not made up of large sheets. And it has some useful properties, and is at the nanoscale, so, in a braod definition, it is a form of nanotech.

Not really. As for whetehr we'll get anything really spectacular, we'll have to wait and see, but there is definite progress on the quantum computer front.

Ah, well as far as I understand it, many of the Singularity proponents are all for strong AI, which is quite another thing.

What, an argument from authority? You know better than that.

Please Register or Log in to view the hidden image!

But more seriously, he has some points, but I think his timetable is too optimistic, and it is possible that not all his technological predictions will come true.

Indeed. This is something that has been frequently ignored, especially by peopel boosting singularity thinking etc.
Heres a very interesting article about that kind of thing and some of the challenges ahead:
http://www.softmachines.org/wordpress/?p=175

Indeed. See the website I linked to above. But on the other hand, more biologically based nanotech will have similar problems and limitations as our current lifeforms on the planet earth have, which will put a crimp into the dingularity freaks hopes.

DO a google search for "vernor vinge" singularity. You'll fined enough to be getting on with.

 

I read 'the singularity is near' a cpl months ago. Very good read.

I find his near term arguments about linear vs exponential growth to be very compelling. He throws in enough charts to make your head spin.

The closer he gets to his proposed singularity and beyond...the wackier he gets. This is however a sort of necessity, as the further into the future we go the harder it is to make any relevant claims....obviously. But once we start talking about wormholes and such...we are firmly in the area of "sci-fi". You will have to read the book to know what I am referring to.

I think there will definately be limiting factors to the exponential growth that Kurzweil does not see. This doesn't mean however that we will never see a 'singularity'. Perhaps even in many of our lifetimes.

PS. I've been meaning to read "Marooned in real time" for awhile now. I need to get to that...

 

I read Kurzweil's book recently, and that's about what he predicts. He projects the singularity for about 2030. He's also hoping for immortality treatments before that (you can tell he really wants to live to see the singularity, but isn't sure he'll make it).

 

is it the appocolypse or something?

should i get in my bomb shelter?,

peace.

 

How would he see it?

 

Dont worry humans wont have to think about things U r worried about.

 

iRobot