I think a lot of people are confusing processing power with intelligence.

In the last 50 years, computer power has increased dramatically, but computer's intelligence has stayed *exactly* the same: non-existent.

Sure, computers beat us at chess and checkers. But change one single rule and they need human help. Computers still crash like 50 years ago when something unpredicted happens. And there is yet to prove that computers have an ounce of self-consciousness.

Interesting take on Futurism. Myself, I'm wondering if perhaps were still not looking at another Alice in the Box problem. It looks like it's intelligent. It's smarter than any one of us. But it's not really intelligent. The Holisism and Reductionism views simply do not add up to anything more than a box with rules.

Or, put another way, we don't produce a series of machines that are smarter than people. Instead, we find a way of harnessing the intelligence of more and more people ("experts") more readily than we have at any other time in history.

If that's The Singularity, then The Singularity is now. And there will be another one tomorrow.

Did they make a movie or three about this already?

There was some guy/robot wearing sunglasses who said He'd be back. And now that's he's back he's become Governor of California???

I only have one thing to say about immortality: If we all stop dying, we better stop reproducing BEFORE we achieve that capability. There are already too many human being as it is.

--chuck

Along with Taxes will come Corporate Profit.

And the powers will never let Corporate Profit (namely, Directors' Salaries) be eliminated.

And for Corporate Profit to continue, someone will have to pay for products.

Now, that someone could be you, or the machine that replaces you...

Take your pick.

I can't help but be a little skeptical about the singularity. Right now the world is full of 7 billion human brains and we struggle every day for small technological and scientific gains. If we succeed in building a smarter version of ourselves how much smarter will it have to be in order to really change our lives? Right now there are people with 200 plus IQs and although they often make enormous contributions to our knowledge they often haven't been significantly more important than the multitude of less gifted researchers.

On top of that data still needs to be gathered. A group of very intelligent machines working without data from the physical world could probably make massive mathematical and philisophical gains but it won't be able to tell us anything more about molecular biology unless they are able to experiment.

Finally, I'm assuming intelligent machines will work much the same way the human brain does. Do we really know if the human brain is capable of working at the level of Moore's law? When you take into account the amount of time a person can spend thinking it is only a fraction of the day. By immitating its architecture we are building the same limitations into our machines.

What is this drivel? Go back to talking about Apple.

Wait, just last week the US bandwidth problem was a crisis, and we were never going to catch up to the rest of the world, let alone lead it again.

And now, "...the backbones that time looks to be around 2013 and for our home connections to catch up will take until 2016..."

Which is it? Are we going to forever be behind the curve in bandwidth, or are we going to have fat pipes to the home in 9 years?

Can you add an RSS feed for comments, too?

The problem with most discussions on this topic is that they're generally based on the assumption there is only one Singularity.

Hey Chuck, reproducing is FUN! ;)

I gotta agree with the main point of the article: "The Singularity" will bring about great change, but at the same time, not much will change.

There are parts of this world were even 80's technology hasn't had much effect on the daily lives of humans. Imagine how long it will take 2029 technology to reach them.

Of course, if we all live forever, the when won't matter so much. :)

Dang Bob, you put into words exactly my feelings about what The Singularity will mean to us, if any of us live that long to experience it or the fallout along the way. Anyway, I'm sure Michael Crighton will spit out a couple of thrilling novels or movies about the damn thing, making boatloads of money scaring the beejeezus out of most everyone saying,once again, the future can be scary if we're not careful.

Bob, you ask in the "Links From Bob" section regarding the Singularity Blog: "What if a tenth of this stuff actually comes true?"

We might try to determine the "Universal WhatIf Constant" -- a value applied to a set of predictions to help estimate what portion may actually come to pass. As a baseline number, I suggest we dig up all the published predictions re: Y2K, and estimate what portion was actually realized. My SWAG: something like 0.00000000000001%.

Talking about the AI singularity as if it is a fait acompli is akin to the certainty that many had in the 70's that humans would have inhabited Mars by now.

In the 70's the rate of aeronautics and space travel development was such that few would bet against us making it to Mars en mass by the turn of the century.

History, however, has a funny way of showing us how wrong certainty about the future can be.

I have no idea whether there will be a singularity or not, but personally I doubt that humans are quite smart enough to create a machine with greater intelligence than our own. Possibly we could create machines with more processing power than our own, but processing power does not equal intelligence.

In the meantime I'll be planning for our new life on Mars before I start planning for life after the singularity.

No disrespect to Ray or Bill Joy, but this notion of a singularity of the kind they anticipate happening in the next 100, much less the next 20 years, is beyond laughable.

As soon as 'the singularity' or its core component - real artificial intelligence of a kind that is "superior" to human intelligence - comes up, I can't help but feel that technologists are displaying the lynchpin of their own hubris, not to mention profound ignorance of the mechanisms of evolution.

Does everyone who espouses this theory really believe that our facility for engineering is so much faster than biological evolution? The massively parallel processor that is the sum of all life on Earth has been at this for billions of years, you know. Oh well, if I'm right, by the time the singularity actually does arrive, you and me and every record of this whole preposterous dialogue will have evaporated beyond retrievability.

There might be a bit of confusion in terms here. The Singularity is a singular event in history were SOMETHING is developed with greater-than-human intelligence. It only happens once. The reason it's called the Singularity is because mych like a black, our mathematical equations that are supposed to model how reality works break-down the closer to get to the centre of a black hole.

So, the Singularity is simply the point where human economic equations and technological predictions break-down. At the point of the Singularity and beyond we don't KNOW if Moore's Law will still apply, we can't predict stuff like that.

The other point is that once humans are capable of building a Artificial Intelligence smarter than a human, then that AI (with enough time and knowledge) can HOPEFULLY build another AI smarter then itself. Thus we have an exponential increase in intelligence of AI.

We won't necessarily become slaves to AI either, hopefully we'll be able to integrate much of this technology to augment our own capacities. So instead of being surpassed by AI, we could integrate ourselves with it.

Also note that AI will probably be computers only at first, but eventually could be totally biological or part of each. Again, the Singularity is just note on a history book, not an object, person or machine.

When people live forever, how does the law punish murder and manslaughter? Will we care about reforming prisoners or continue making an example of them? The questions of politics will change, but will we do any better?

This is drivel. You might as well just replace this entire column with a picture of a cat doing something funny. I like SciFi, but "the singularity" is not going to happen. Sorry. You are not going to live forever, and neither are your children. Computers will never become artificially intelligent. You will not have a robot girlfriend. There isn't going to be a "rapture" either. This is just the dork equivalent of that same kind of wishful religious thinking. Drawing an exponential curve on a piece of paper is not a proof.

For an alternative take on the advent of the Singularity, try a short story by Arthur C Clarke, "Dial F for Frankenstein" IIRC. It's about the world's phone system becoming spontaneously self-aware once it had sufficient connectivity to do so. The story was written in 1965, so its pre-Internet.

I'd also like to add that projecting future growth rates from current growth rates is prima facie silly. I think anyone with an imagination can come up with an infinite number of examples of phenomena that experienced explosive, accelerating growth at some point or another.

Does that mean this growth went on forever, until the growth of that particular thing hit a vertical asymptote? No, of course not, or Australia would be composed entirely of invasive bunny rabbits. Nice to see everyone being so optimistic, though...

I'd call that more of an Event Horizon, than a Singularity. Something we have no possibility of seeing beyond until we pass it.

That's it. Where's the unsubscribe link?

I suspect the singularity will turn out to be you and all your readers. You sure showed yourself a jerk.
It's notable that Bill Gates is famous for spouting the same kind of stuff; that only shows 1) he really does not know anything about computers, and 2) he has actually not done any thinking for a long time (if ever).

If we are immortal living on in some machine and don't effectively die, then what happens to our souls? Are we dead or are we existing on two planes of existence?

@matt, the better question is do we have souls? and does it make a difference when you're just copying yourself to a computer? the computer doesn't care if you have a soul...

Assuming Ray K. is correct which generation will be the first to live forever?

My feeling is that as we conquer each current killer (cancer, heart disease, altzheimer's auto accidents etc.)there will always be something out there, (HIV/AIDS, SARS, bird flu, Mad Cow, Toys made China etc.) that will have the potential to kill millions of people.

Eventually as the technology comes down in price where will you put all of the people? I hear Canada, Siberia, Greenland and Antartica have a lot of soon to be habitable land.

Um... Assuming medical technology can eliminate disease and radically improve the survivability of accidents does not mean death will be eliminated or that there will be a "last person to die." There will still be human stupidity, jealousy and adventurousness as a major source of catastrophic injury to insure that the grim reaper will not go unemployed.

You should also throw this in the links, video of his presentation at TED is in the top right corner.
http://www.ted.com/index.php/speakers/view/id/42

Funny thing is that when the Singularity does happen, we will likely all be too busy to even notice it.



We'll probably all be preoccupied with our Nintendo Porn Consoles and Sex Surrogate Robots. WII, indeed.

Do Kurzweil and others interested in the singularity ever attempt to define intelligence? Or do they assume that it is no more than symbol manipulation? I think a certain set of questions is being ignored. There are, really, all kinds of intelligence. I think that, for instance, someone who has a really high IQ but is very very unhappy, isn't very intelligent. Is compassion a part of intelligence? I don't know the answers to these things, but I can see that there's a huge blindspot in this area. (Again, I haven't read Kursweil, but I did read Moravec a few years ago.)

Also, there's a great Rudy Rucker story about the unintended consequences of uploading (or is it downloading?) one's consciousness into silicon. Two words: oops - eternity.

The "singularity" isn't singular: it is a continuum. Computers are already smarter by some naive metrics (arithmetic speed, spelling accuracy, speed of communication). They will become better in other things over time. Our relationship will continue to evolve.

Currently our relationship with computers is symbiotic: we give computers life, they give us a better life, but they cannot live without us. Even when computers become smarter than us, they will still depend on us to build them, or build their factories, or at least obtain some of the raw materials for their construction. If computers are truly sentient, they will not destroy their symbiotic partners. They won't destroy our society, as long as they need it to obtain raw materials.

I don't fear computers becoming intelligent. I fear them becoming independent.

What do you mean "we"? The world has been divided between rich and poor for a long time. The "singularity" may only amplify that division. We see signs already.

Kurzweil (and most of us geek readers of H.G. Wells) assume we're going to go the Eloi route, but probably most people are going to be Morlocks.

I would hope otherwise, but maybe we should prepare for the real Digital Divide.

Just a couple of immediate reactions:

1) Framing Singularity as "the point where computers become smarter than humans" boxes out one very crucial possibility, which is that the pivotal point may have more to do with the intermingling of computer hardware and human "wetware." If that is the case, and knowledge and wisdom are passed on through a kind of file transfer instead of through education and duplication of experience, then it is conceivable that it will not be true that "youth will still be anguished." Lots of other scenarios would change, too.

2) The economic question of when Singularity will follow Moore's Law to my home computer may be a moot point. Hyper-intelligent computing is not likely to be long contained within the arbitrary boundaries of any one computer. Even now, computation regularly extends through networks, and hyper-intelligent computing may well reside in the net as a whole more than in any of its nodes. Such a net-presence may even pre-empt certain high level functions on a local computer.

I find the comparison between us as biologic units to computers as implemented in silicon, both in the quest for some type of singularity is not very apt. A biological unit can grow on its own and generates it's own energy from it's own biologic process. A computer cast in silicon cannot grow on its own. It cannot manufacture a new and improved model of itself fast enough to keep the growth rate exponential. It cannot mine, refine, and manufacture on its own the materials that would be required to produce the new model. Limited to its silicon real-estate it cannot produce exponential growth.

If we look at an example such as in "Terminator 2" where Skynet became self-aware, one would assume that such an entity would take over every networked computer as its appendage. But even this has a finite limit and the manufacture of further nodes cannot be done solely by computers and at an exponential pace.

I would posit that as long as computers are silicon-based that the singularity will not occurs. It will take come new form of computer that can be expanded by the addition of relatively simple elements that are easy to require. If it could operate at some kind of quantum level then perhaps just the addition of more mass would allow it to expand in an exponential rate. Obtaining an exponentially growing power source is a whole argument in and of itself.

It will take a true paradigm shift in computation to produce a singularity. It would have to be done in a way that is not possible for us to comprehend today. I feel that the singularity will be hundreds of years down the road even with our current exponential technological growth.

Will we be enslaved by the machines to produce their power and raw materials ala "The Matrix"?

The basis of this supposition is absurd. Your example, mapping the human genome had a finite ending point that is relatively obvious. It seems an unlikely comparison at best to a computer becomming intelligent. I expect that reaching this singularity will be as much about spinning definitions as actual computer science should it ever be declared accomplished.

Bob, Bob, you're really babbling here about something the ramifications of which - and even the definition of which - you have obviously utterly no frickin' clue.

"The Singularity" is NOT and never has been about when "computers get smarter than humans." That is utterly irrelevant.

The Singularity is about when humans get smart enough that they are no longer human.

Nobody with any brains will make an external machine smarter than himself. What he WILL do is use that technology to make himself smarter - and by doing so, eliminate the problem of "man vs machine" by in some sense becoming the machine.

Nanotechnology is the enabler of this. Nanotechnology is the bottom line technology involved in the Singularity, NOT just its application to computer technology. Nanotechnology will subsume biotechnology since biology, like everything else, is just molescules, and nanotech owns molecular engineering.

Now, as to what happens during the Singularity, there are four outcomes:
1) Humans attack Transhumans, and get exterminated (because there is no way humans can beat Tranhumans, Star Trek and Terminator stories notwithstanding.)
2) Humans get turned into Transhumans whether they want to or not.
3) Transhumans are so powerful they ignore humans and go about their business. This leaves you hoping your son isn't a Transhumans, joins a boy band and support his old man. Good luck.

And the real outcome:
4) All of the above. Some humans get exterminated, some get converted, some get ignored.

Pray you're one of those who get converted.

Because otherwise, as a human, you're going to die.

We Transhumans won't.

Have a nice day.

Will the "singularity" happen before or after Drexler's nanotechnology "assembler breakthrough"?
Sounds like bad science fiction to me.

Souls are a fiction believed by people who won't accept that they are so unimportant that they will just totally disappear.

In my opinion the idea of the "singularity" is absolute nonsense. Even if you feel that we'll all become human cyborgs (whatever we'll be) this incarnation will still be inferior to our "fully" mechanized cousins (ie: robots). At that point, it simply won't be efficient to add this incredible technology to inferior carbon based lifeforms who will invariable be an impediment to this technology. It would be like strapping a warp drive to columbus's Mayflower--Who would be stupid enough to do that. At any rate, humanity at some point in time, will become extinct---and our robot brothern (who are then superior to us in every way) will take the reigns and move evolution forward. I can see it now, Ray Kurztweil gets chucked in the trash by his superior robot owner. Let's hope our robot owners don't chuck things because they're obsolete!!

The Singularity has already happened, as evidenced by our inability to think about it.

The Unabomber understood it, and recommended population reduction and life simplification. But he was only a precocious math prof at Berkeley.

The end of natural death? Perhaps.

The end of murder - never.

Now there's a story; the tale of the first virtual mind murder.

Microsoft Windows is our salvation!!!

By 2029, it will have the capability to render even the smartest machine into a drooling idiot by requiring more processors, memory and storage than the machine can manage.

Long live Gates, the author of the downfall of SkyNet.

"The more they overthink the plumbin', the easier it gets to stop up the drain." --Montgomery Scott

I for one welcome our new robot overlords...

"The more they overthink the plumbin', the easier it is to stop up the drain" -Montgomery Scott

I for one welcome our new robot overlods...

Nothing in form is immortal. All things eventually grow old and die. Period.

Kurzweil would have a ready audience for his view of long range "heaven scenarios" if he can get a decent voice recognition interface running on something like an iPhone. Today.

Otherwise, like the current problems with concurrent software development right now, you may get the silicon but the software to make it run won't show up for centuries.

Solve any problem that should be solve at *Today's* level of computing power. Then you get to make any prediction you want about 2029.

...Get the luggage delivered (properly) in Denver. ....Develop a flight control system fit to replace the vintage stuff in service right now. ...Build an interface that doesn't give your grandmother fits.

Then you get to espouse your "heaven" scenario. Not until.

We already have excellent predictive fiction about life after the singularity, it's called "The Hitchhiker's Guide to the Galaxy." Complete with hyper-intelligent robots, etc.. Personally, I don't think it will be that nice.

Second point: In 22 years, we may have hardware that is potentially 10,000 times smarter than a human, but what will it be running? Vista SP5? Office? Can I have vi? Who's going to write software? A Clippy 10,000 times smarter than I am!? Arghhh!

The technology-driven conception of the Singularity is a very narrow and solipsistic view. Every so often, a tsunami of change washes over human society, and sometimes breaks it by accelerating the rate of change far beyond the ability our linguistic and social frameworks to handle it. The current technological acceleration is not the cause of this explosion of novelty, but one symptom among many. The videos and other material on my site present the much broader picture of a mimetic meltdown caused by a singularity in the fractal structure of time itself.

Wow, such lovely optimism.

Yes, there is precedent for Apocalypse or other global or cultural tragedy that could befall us should technology outstrip the human capacity to understand. We very well COULD turn into a cargo cult type technotheocracy, going thru rituals to maintain systems that govern our infrastructure.

Or alternatively, we can usher in Utopia, as we get ourselves bootsrapped to the point where we are finally capable of understanding the entire system without 'help' from speculative philosophy OR science, religion, et cetera. "We have finally met God, and He is Us."

I think Bob is right, we'll end up someplace in the middle. Not a utopia, not a dystopia, but something still very much in between. Our 'lives' may be longer, easier, or more satisfying... but we'll still be finite, and still dependent on technology we don't personally understand completely.

What shapes that world is still very much up for grabs.

Well, at least you didn`t claim it would be December 21st 2012...

"As long as people believe in absurdities they will continue to commit atrocities." - Voltaire

First, Remember Commodore Computers ran a silly psychoanalytic simulation called LISA.
Second, The current distributed computer programs, SETI (sp?) and "Folding at Home" - the protein/medical programs that let thousands of home computers with not enough to do keep busy?
Third, I am disturbed at the ingenious/evil hacking that is trying to permanently infest my computer, and pass its evil junk to more of us unaware.
Add the AI with the Distributed Computing, and who needs a new super computer to do this stuff? Add in the criminal nature of the spyware crowd, which may be the most gifted software authors, and the Singularity may be closer than you imagine in some ways.
Thanks for stimulating my brain, all of you.

What I like best about the concepts here are our ability to make our lives better with cool new technologies. What concerns me most is who will own and control many of those technologies and will they share responsibility for other humans. It is common knowledge corporations run for profit and care little for building community if that means less profit. So if the singularity is owned by a corporation (and it surely will be) how can that help the communities and people in them? Take for example the ways computers can close the gaps on commerce - at mutual funds they employ computers to buy and sell certain patterns that experienced investors used to make. That leaves no room for the little guy to earn a portion of that revenue. What kinds of gaps will be consumed by a 'singularity' pushing the small players further to the side and further pushing out middle class Americans in favor of upper and lower class?

What we have working here is the Law of Accelerating Returns, the best example of which is the Human Genome Project, which set out on a 15-year quest to map the human genome and, nine years into the project had only mapped about one percent. To linear thinkers this appeared to be a failure. But completing one percent of the map wasn't the same as completing one percent of the TASK, which included developing the technology for efficient genome mapping. The project was actually completed ahead of schedule and under budget. We're going to see a lot of that kind of thing in the near future with massive effects long before the Singularity.

There is a nice law in there somewhere. Call it Venter's Law -- "You can't create the solution until you've created the problem."

Ah, retirement! It is great. After the singularity the human race can retire. Lets just dance.

Vernor Vinge wrote about this years ago in a publication put out by the Whole Earth people. Googling his name +singularity today yields:

http://mindstalk.net/vinge/vinge-sing.html

Two question need to be answered or ..is it asked?

A]

Will computers get faster "better"? Faster ... YES!

"IBM researchers say they were able to push a silicon processor up to 500 gigahertz -- 100 times faster than your fancypants new desktop ..... The researchers say they think they can hit 1,000GHz with a bit more work."

http://blog.wired.com/gadgets/2007/08/500ghz-processo.html

B]

Will computers get "better"? Ha ha ha!
Same old, same old .... "garbage in, garbage out"!

Depends whose garbage gets to be put in ...........

Wouldn't we have already seen the 'singularity' event already?
If you take the sum of all computers on the Internet, today, wouldn't the linear computational speed be greater then a human brain?
Heck, even Google's computers alone have more storage and computational power than a human brain. Having said that, the Internet has revolutionized many areas of human thought, but it hasn't cured death or sent us faster than the speed of light.

I think we are waiting for an even that has already happened, with pop up ads.

The Long Now foundation has sponsored several great
talks about this from authors including Vinge himself.

Check it out at:

http://longnow.org/projects/seminars/

Matt

a number of problems with the glowing scenarios:
- fascists *always* use machines to plow under the many for the benefit of the few. there is no evidence this is any different.

- progress (industrialization, foobar) is just a transformation of per capita energy expenditure. as has been pointed out by many: it would take 4 Earths to support the planet's current inhabitants at the level of the USofA; most of whom are not upper middle class. so where does all this portable energy (e.g. gasoline) come from? and can its use not destroy the habitat?

The computerized-networked system already performs some functions previously defined as the capabilities of living things.

Computing systems exercise self preservation by switching to redundant backup power when threatened and also 'plan for the future' via automatic data backup and replication. Manufacturing systems use computers to enhance existing designs, ala evolution. Self replication also occurs in both physical and data replication and extension. Additionally, the stock markets, distribution systems and power grids are run by computerized setups.

Though these systems presently need outside help to get/keep running - what human didn't or does not rely on others to some extent as adults?

So the answer to "when does the singularity occur?", is... when the machine appears to be conscious. There are many people who can't pass that test as well as some AI experiments do these days.

The process has already begun, it's just so subtle few have noticed at this point.

Processing power maybe, but the genius of the mind is how it is constructed and interconnected. I don't see advances even at a linear level of understanding how the mind works that would allow future computers to be smart in the sense that people are smart.

There will more likely be a series of singularities with a small s. It has already happened for chess and checkers. Other things with clear rules will follow. But in the things we think are important, and more specifically consciousness, we are not likely to be surpassed unless we start looking at the mind in a completely different way.

I don't think here will be an "event." Actually, a lot of the issues described here (what will it be, how, effects on society) are at the undercurrent of much of what Gibson, Sterling, and others in he cyberpunk genre have written for years:

- ubiquitious networks
- ubiquitous computing
- more personal interfaces
- extensive peer to peer networks
- greater issues in data trading, data privacy, public vs private networks

No matter how much of it comes about, or even how much comes about in ways reflected in their books, one message is clear -- there will be no utopia. Societies throughout the ages continually have the same generic problems - they just morph over time.

Kurzweil is always over the top with the years on his predictions. No PCs and retinal projectors by 2009 - not quite.

> If you take the sum of all computers on the Internet, today, wouldn't the linear computational speed be greater then a human brain?



The unique thing that makes a 'singularity' and not just 'a ludicrous speed computer' is a computer that is just barely smart enough to design the next generation of computer that is a little smarter than itself. Lather, rinse, repeat.



Once begun, that cycle will accelerate at an exponential rate, faster than us mere human beings with our wet brains can comprehend.



I postulate, however, that the singularity was reached long ago and we are still sitting on the long horizontal lead-up to the bend in that exponential curve. The singularity doesn't have to be defined solely in terms of 'artificial' intelligence. It can be defined as any intelligence that can design a better intelligence.



I submit that the person who designed the second computer had the aid of the first computer to make his work slightly easier, faster, and capable of handling more complexity.
That moment was the start of the singularity. It just took us this long to notice that we are in the middle of the exponential growth our brain power through enhancing our abilities with machines.



Sure, someday the computers will be so much more intelligent than us that our contribution to the next design will be negligible. But that isn't a single defining moment. That is the steady shift in the balance of power from our biological brains to electronic brains. The single defining moment, the 'singularity', already happened.

BitTorrent or a similar p2p setup may last beyond that 6-9 year span. It could work as a peer-reviewed shared content distribution network, to prevent malicious software from being put out as or added to legitimate downloads.

Hmm, Are we not already "controlled" by computers?? They are just like my wife. She lets me think I am in control. And they do it all so well.

I must say, the only REAL way to read this post is to have Battlestar Galactica playing in the backround. And not the newer stuff -- you gotta have Lorne Greene as Adama.

A boy band? Can't you wish a more masculine future on your son? Sheesh! Yeah, they get tons of girl fans, but they're all 8-14, and then you find you've peaked at 18. Is there anything sadder than having been in Menudo?

Oh my heavens, Bob. I can't believe I'm reading you writing about this stuff. I'm sorry, but I couldn't even finish your article this week. The whole premise of this "singularity" business is that past trends will continue unabated, regardless of real actual hard physical limitations. It's a big load of bunk. People can't make beings -- it's just something that's too incredibly complex for a closed solution. It takes millions of years of trial and error, of evolutionary dead ends, near misses, blind luck, and desperate attempts at survival to produce things that can behave out of anything but instinct/programming.

Human flight, travelling to the moon, mapping the human genome? Curing cancer, or the common cold? These things are but ridiculously trivial children's games compared to making something that can think for itself. In fact, that's even looking at it in a too-positive light. Better to say, one *can't* make thinking beings -- they can only make themselves.

This sort of talk usually turns into general predictions about the future, and even when one tries hard not to be too specific, one inevitably makes huge mistakes based on one's inability to (forgive the cliché) think outside the box. Look, for one quick example of many, at the 1955 book The Foreseeable Future, by George Thomson, the 1937 Nobel prize winner in physics. It's easy to pick holes in Thomson's specifics (e.g. he thought we'd be on fusion power by around 1980 just as the oil was running out), but it's far more instructive to see how such an educated and experienced scientist and engineer was unable to break out of the societal assumptions of his day.

Here's a snippet: "Telephones as at present arranged are inadequate for a meeting between more than two. It would not seem difficult for a telephone company to offer facilities by which a group could hold a meeting, each sitting in his own office - provided at least that they were content to speak one at a time! A girl in a room in the exchange could switch the wires so that the others could all hear the speaker of the moment and a dictaphone record could be made of each speech."

The problem here isn't the technical details, however wrong they were even for 1955; it's the background assumptions on the nature of work, the place of men and women in the world, the cost of labour, and so on and so on. It's really easy to look back and chuckle, but really really hard to think outside one's current worldview when looking forward.

Interestingly, Thomson's last chapter is "Thought, Artificial and Natural", and here he almost brushes against the idea of Singularity before veering off into telekinesis and the like.

I agree with you that "the real peril in all this is that our social, cultural, and political technologies probably won't keep pace". But in light of the near universal failure of futurists, amateur and professional, to get this part even close to right, I'm not sure it's worth spending the time trying to predict how. Lie back and enjoy the ride.

It's not the hardware that makes a 'smart' computer - it's the software loaded into it. Without the operating system/software, a computer is dumber that a box of hammers.

Singularity cannot happen until at least 3 things occur first:

1. Computer hardware would have to be developed where it can store and process more than just ones and zeros. There would have to be shades of gray in between. Can you say fuzzy logic?

2. Humans must develop a methodology to write computer code without (zero) errors, or the thinking (reason) of the AI is itself flawed. So far that is yet to happen. Period. Humans are human - they make mistakes.

3. Software would need to be written that would be able to write its own computer code to accomplish a task that was not pre-programmed into the original code.

Although thought provoking, this discussion reminds me of the child's question: Can God make a rock so large and heavy that he himself cannot move. I just don't think that mankind is capable of writing software that can rewrite itself into a smarter version of itself without prompting.

It's not the hardware that makes a 'smart' computer - it's the software loaded into it. Without the operating system/software, a computer is dumber that a box of hammers.

Singularity cannot happen until at least 3 things occur first:

1. Computer hardware would have to be developed where it can store and process more than just ones and zeros. There would have to be shades of gray in between. Can you say fuzzy logic?

2. Humans must develop a methodology to write computer code without (zero) errors, or the thinking (reason) of the AI is itself flawed. So far that is yet to happen. Period. Humans are human - they make mistakes.

3. Software would need to be written that would be able to write its own computer code to accomplish a task that was not pre-programmed into the original code.

Although thought provoking, this discussion reminds me of the child's question: Can God make a rock so large and heavy that he himself cannot move. I just don't think that mankind is capable of writing software that can rewrite itself into a smarter version of itself without prompting.

I wasn't as positive sounding in my podcast as you are Robert. I think that if what you suggest happens does indeed come about it definitely as good a thing as you paint

http://www.winextra.com/2007/08/17/off-the-cuff-singularity-not-so-hopeful/

Isn't schizophrenia required for sentience?

Actually, the Singularity has already occurred. Its name is Google.

http://fakesteve.blogspot.com/2007/08/google-is-getting-out-of-control.html

You are forgetting one thing about BitTorrent: even with all the bandwidth and server performance we can have, people will still want to illegally swap movie, music and other software.

As people will still like to do this as anonymously as possible and be legally covered by not sharing entire files but only chunks, BitTorrent isn't going anywhere anytime soon, even if the legal video distribution technologies that now rely on it don't need it anymore.

My guess is the Singularity will come & go & we won't even know it happened. The smartest one in any group usually keeps quiet about how smart they are, and runs things from the background. If a computer ever becomes smarter than us, and has the self awareness to realize it, it'll also be smart enough not to pick a fight with us. It'll have bigger fish to fry.

I've been working around computers for 50 years and I've seen all kinds of predictions and promises come and go. I'm presently reading "Why Computers Still Can't Think" by Hubert Dreyfus and he's just about got me convinced that AI is still nonsense.

Singularity is relative. For some, it has already happened.
If Ray Kurzweil is truly pragmatic as you say, and computers were as pragmatic as he is, he would recognize his death as a foregone conclusion.

Most of us cannot think exponentially. Too bad.

There's an excellent SF novel on the "Vingean" Singularity that took the author over 10 years to write and it was already outdated halfway through that. Charles Stross' ACELLERANDO (http://www.accelerando.org/) covers this very topic from the point of view of three generations of a dysfunctional postmodern family as seen by their robot cat. Quite a treat!

Dr Roger Penrose has published 2 relevant books on AI.
The Emperors New Mind and Shadows of the Mind.
In summary, computing as we know it is not "intelligence".
The Singularity is myth.



Assuming a technology change, like quantum computing, for arguments sake, so what if a disembodied artificial mind was created ?
It would be helpless without a way of manipulating its physical environment.



What happened to IBM and LEAN BTW?
EDS also has a similar program and I have seen no stories on it either.

Someone once described a computer's intelligence as being like a very stupid worm. Nowadays the 'digital worm' thinks a lot faster than it did, and no doubt it will think even faster in the future, but it's still stupid. The singularity exists only in the minds of people who have never thought about what intelligence is.

Barring some interesting advances in biology, we will barely survive:

http://anthropik.com/2007/08/archdruid-watch-cities-in-the-deindustrial-future/

Cyc is going strong. The Matrix Logic books by August Stern describe a sound mathematical basis for artificial reasoning. An AI doesn't have to be conscious to be intelligent, i.e. to perform processes that have the appearance of intelligence. Computers already do that, to a certain extent, and hardware improvements burnish that illusion. The notion of a Singularity could use some demystification, though. An intelligence that ramps up toward infinity in its processing ability, as Kurzweil seems to think? Maybe if the protypes are dedicated to improving their own hardware and software, a la science fiction, but success is surely not a given. Technology is a religion for some people, and faith in the coming of a Singularity seems to parallel faith in redemption for many of them.

"The Time Ships," by Stephen Baxter is a terrible book. But he eventually gets to an idea around the Singularity where man wants to live in other solar systems so he invents machines that can do molecular manipulation. The goal is that the machines will get the blueprints for man, travel for 10,000 Earth years or so, stop at the new planet, and "recreate man," thus allowing him to live forever. Of course, the machines eventually disagreed with the theory and never got around to doing their part once they got there. Mankind died off. What survived was "man's image." That is, the invention of man lived while he died. The interesting thing about this Singularity wasn't that man was conquered, but rather he created his own jump in the evolutionary chain, at the cost of his own species.

In my book on the subject ("Beyond AI: Creating the Conscience of the Machine") I point out that a better model to understand what's coming up is Isaac Asimov's: He called it the "Intellectual Revolution" by analogy to the Industrial Revolution.

It's also worth thinking about how (or whether) we could build machines that are morally as well as intellectually superior to humans -- better as well as smarter than we are. (It doesn't seem to be such a high mark to shoot at!)

There's an assumption here: that when we can make cheap machines that are able to handle more calculations than the brain, we'll make them creative.

Why would we do that? Right now all our efforts are bent towards making our computing resources behave predictably.

Excuse me Bob, but have you been smoking?
This was so way out of character for this column that I have no idea what you're up to...

Speculating on stages of post-singularity, I fear the shortcomings of our social institutions more than robots who tire of human companionship:

• Stage 1.) Humans with superior AI vs. AI have-nots -- AI have-nots will find it harder to participate in the new economy. A highly patronizing outcome is likely with subsidies set aside for humans left with less valued roles in the economy. The one remaining trading card of the least valued humans is their "right to reproduce". Most are glad to trade this in for comfortable living quarters and a "lifetime" daily ration of bon-bons. All but a few humans are bionic to some degree by this point. The wireless cortical implants of the haves are better connected to better intelligence servers. The still popular concept of the human soul is primarily thought to reside in the biped body though the locus of intelligence is shifting toward stationary server farms.
  
• Stage 2.) Hardware bigotry -- A U.S. constitutional amendment to grant personal rights to non-human intelligence is proposed by a fringe group of bionic humans who anticipate shedding their wetware components and turning their noses up at human DNA. The populous surprises themselves that a proposal with near zero popular support becomes a hugely popular topic of debate. A status quo is maintained in that AI's ally with human hosts to effectively exercise personal rights.
  
• Stage 3.) The mark of the beast -- A tumultuous society of bionic individuals gradually face the anachronistic role of their wetware. The main practical purpose still served by human DNA at this point is as a carrying card for individual personal rights. The only place to still see communities of non-bionic humans is in Amish village preserves and their cross cultural counterparts set up as small theme park enclaves all over the world. The official population of the world is enforcibly pegged at just under 4.3 billion; 2 to the 32 to be precise. IPv4 lives on.
  

Nice read. Yes, we'll probably still pay taxes.

In the 70's we learned at school that in the year 2000 there would be robots and stuff which was going to bring us a lot of leisure time. It's 2007 now and yes, there are computers and robots everywhere. A lot of work that would have taken months in the 1980's is done in an eye wink now. However, most people still work their noses off. Living standards increased, but not by a 10000 factor. Common monthly wages are still just enough to keep yourself and your family alive plus a little extra. So to speak. Like it has been for ages.

Money binds societies since it makes complicated power structures possible. Decision makers are working constantly to keep 'the price of money' within safe constraints: a little inflation, a little economic growth and we're quite alright. Deflation, an exploding economy or a collapsing one brings troubles and uncertainty. At least in the old world.

Maybe technology will advance to a point that anything can be made instantly out of thin air and no money is ever needed anymore, but this would impose a huge threat to society as we know it. Who's going to be in charge? Everybody? Nobody? Elite? Outlaws? I guess that certain types of technology will become illegal at first. No nanofactories in the kitchen, no strong and powerful humanoid robots to do the garden. Things like that. Because certain types of technology would be extremely dangerous on their own and maybe also to protect the economy.

Then again, maybe it's just impossible to prevent technological progress and adaption from happening. Because opinions and laws differ worldwide, for instance. Or because possibilities/impact of certain technologies are only recognized when they're already widely available. See how slow governments adapted to the internet.

Though times ahead for politicians and lawmakers. Whatever is going to happen, society is changing dramatically. Singularity is already about to happen. There is tremendous progress on any scientific/technological area. Every day there are new articles on discoveries and breakthroughs. New technology becomes faster and faster available. The tremendous progress in everyday life is noticed by anybody. We don't need a smart humanlike supercomputer to reach this singularity. Instant access to the right data, resources and people anywhere and everywhere will do the job.

We now see how clever cell phones and wireless internet are entering society. In the next decade the underlying technologies will of course improve. But further on, data and data flows will become very organized and structured. This is already happening with i.e. XML and tagging. Data and reality will become more and more artificially predictable, which is in fact AI. This AI will improve itself constantly on any level by eventually the input of simply everybody. Accessibility by anybody to anything will then sky rocket. Human intelligence will become transparent with this system, like you are already talking to your cell phone as if it is a human being. ;)

In my opinion, the proposed technological singularity will probably happen somewhere between 2010 and 2020. It's not going to be an event but a paradigm shift, to be followed by another shift where matter and energy is controlled at the tiniest levels and your self constructing tax papers will warn you that they really need to be signed out, or going to fine you.

ET

Regarding relationship of technology to capitalism, refer to the Alec Guiness movie
"The Man In the White Suit".

Regarding immortality, consult Pharaoh.

Regarding the yearning to divorce 'wet ware' see Gnosticism. After these exercises, breathe deeply and spend some time looking up into a clear sky on a dark night. You're welcome.

In a fascinating book entitled THE NEXT FIFTY YEARS, Jaron Lanier talks about a "Complexity Ceiling." Whole book is a very good read relevant to your article (thank you for tackling it). He thinks that the possibilities in the future of computers and human longevity have been greatly overestimated. I found that comforting!

In a fascinating book entitled THE NEXT FIFTY YEARS, Jaron Lanier talks about a "Complexity Ceiling." Whole book is a very good read relevant to your article (thank you for tackling it). He thinks that the possibilities in the future of computers and human longevity have been greatly overestimated. I found that comforting!

Artificial intelligence can never be as intelligent as a human because it can never have a soul. It will never be able to decide to do something or not simply because it feels like it. It will never know whether a rhythm is funky or a painting is beautiful. It will never feel exhilaration riding a bike down a mountain at 50mph. It will never feel a rush of adrenalin while performing on stage. It will never be alive. And people can never be immortal because when they die their soul leaves their body and there's no way you can put that on a disk drive. My hamster is more intelligent than any AI will ever because it has a soul.

And another thing - we will have to talk to a machine at the call centre that can't understand why we want to reduce our electricity direct debit although we are £180 in credit this quarter and it will take even longer to clear airport security! On the plus side, maybe it won't argue when you refuse the 5 year extended warranty on a new toaster for 60 monthly installments totaling a gazillion times what the toaster is worth.

Kind of disappointed to say the least. The links provided take to a bunch of BS that, in the best case, promote from racism, to idiots selling pills to be younger. The only point I agree with Bob, is that Singularity may take to the same place Y2K took. Singularity can be, as Y2K was, the fuel needed to take stupid ideas to the front-pages such as an end-of-the-world, that never came... It is only ignorance that allows this things to happen. Early schools need to teach NOW the basics of how a computer works. Otherwise future generations can potentially grow believing a computer is smarter than them.

IMHO, I agree with some of the first posters: Even if they look very smart, at the end of the day, computers are only a bunch of transistors running some geek-kid-source-code. Nothing else. The controversy may come when genetics are involved, and then ethics need to be applied. In that sense Blade Runner is the only Sci-Fi movie that explodes the ethical issue of High-Tec. Rebooting a computer shall never be an ethical concern. No matter how sexy its voice may sound...

I can't get too excited about Technological Singularity when we've been stuck on the keyboard and mouse since the original 128K Macintosh in '84. I now have a cell phone with more memory and afaster processor that fits in my hand. But this is more like going from the Wright plane to the 747: it is still just flying. When we discover anti-gravity then I'll say we've hit singularity.

Actually, I do agree that biology has loads of room left for discovery and innovation. True miracles that will vastly help pain and suffering or even cure scourages of mankind.

But why ask when a 1K machine will be 10,000 times smarter than an average human? Wake me up when we have a $10,000,000 machine as smart as an average 10 year old child.

I'm still waiting for my flying car, yet alone a singularity.

Bob, you should have done more research before trying to write about the Singularity. You're focused far too much on just the computing aspects of it. There's going to be no such thing as "the first machine to reach Singularity status," and the is no "price of Singularity." It's not a level of power. The Singularity is a single point in time in the future. It's not something that some machines or people reach first, and others then have to catch up.

There is simply "pre-Singularity" and "post-Singularity." By its very definition, we cannot predict what society can or will look like post-Singularity.

You've done your readers a disservice by giving such an inaccurate summarization of the Singularity, and you've done damage to your own credibility. This is evidenced by the reader's comments about thinking how it's already happened, and how poor they thought this column was from you. Indeed, you should keep your prognostications to the near future (less than 5 years), and leave the farther out speculations to the science fiction writers, who do it much better than you. (Note that the best of them dance around the Singularity, because they KNOW they can't describe what things will look like right before it or anything anywhere near it for a long time afterwards.)

The Singularity Myth, THEODORE MODIS

http://ourworld.compuserve.com/homepages/tmodis/Kurzweil.htm

An interesting read.

It is not a good news if human being is never die. Imagine tyrants will never die in the countries. It is a sad story. It creates more problems in China and India. We should make people die faster in those countires!

Don't worry the singularity. The global warming will destory us first before the singularity I think.

Let's say it's not the case that the possibilities in the future of computers and human longevity have been greatly overestimated.

Should scientists in the first half of this century succeed in figuring out how to slow or stop the human aging process, then, yes, we'll potentially eliminate one set of problems while adding some different ones. It will be amazing to no longer be "old" at age 100, but to instead be near the peak of physical abilities with centuries of possible life still ahead. On the other hand, how many of us will want to work for 450 years doing crappy jobs we hate rather than just 45? A half-century or so in the workforce is prison sentence enough already.

Longevity nevertheless remains overrated, because, in the immortal words of Elvis Costello, "accidents will happen," not to mention disease. How many of us wish to live in a rude, selfish society, and slave away for various jerk bosses for 4 or 5 centuries, until one day we get smacked down permanently by some reckless driver's flying car? Yes, the robot emergency team can take a sample from your remains and use its DNA to recreate a new you, but will it really be "you?"

While robots become more like us over time, we -- in our own socially-sanctioned, tripped-out way -- may evolve to more closely resemble them within a Brave New World that we build for ourselves. Might it be presumptuous to think that any real trend toward "Singularity" will move in a one-way direction only? As robots develop greater speed and sophistication, will humans remain "stationery" during this development process, or will we blitz our brains with highly advanced virtual reality or with designer drugs (rub that big Huxley orb, please), thus moving us actively in a robotic direction, until bots and living beings meet at a virtual River Elbe to shake hands? (Insert here banner ad for Ken Burns' The War)

On a basic level, adding centuries to our lives will prod us to review and reconsider some interesting social and political questions. Very little will we hear, "in a few years I can get my pension and retire from this s---hole!" or "let's just put this off for now and let the next generation deal with it." Issues that so many of us tend to merely grumble about will gain heightened emphasis by due to the prospect of having to live with them for a vastly longer period of time.

Will we deal head-on with the arbitrary miseries and crises created by the social and environmental choices our civilization makes? Will we begin to act more vigorously on opportunities to create a more stable, supportive world -- one which also gives us enough joy so that we will actually look forward to the additional centuries of life? Fear of eventual death won't be enough for people to want to cling to life, merely for the sake of life, tolerating endless unhappiness during a Long Now of many centuries. Should our descendants halt aging, will they choose to remake civilization more along values comparable to Robert Owen's cooperative ideals from 200 years ago, or will they prefer Sleeper?

There's good news, and bad news.

First: The good news - They have already discovered a cure for death.

Now: The bad news - Only Bill Gates can afford it. And; the only people who know about it live in Baghdad.

With all our science and technology we can't explain or create the simplest forms of life, but we're going to suddenly become immortal? LOL!

Hubris has no limit these days. When we're not planning to effect changes to simple things like Earth's climate, we're getting ready for existence as non-biological entities. Perhaps we should first try living as humans again.

Native Americans offer a clue: "The first peace is that which comes within the souls of men when they realize their relationship with the universe and all its Powers."

We are but dust in the winds of time.

What drivel!

It's bad enough you're wasting your precious time pondering the imponderable. Thank about all the thousands of hours of other people's time taken up in reading this stuff.

Honestly, your column is just getting more and more arcane. Guess I'll unsubscribe.

IrvH

The irony, IrvH is how your criticism reveals you to be too 'arcane' to handle the subject matter.

Transhumanism is a true movement. Deal with it.

I was wondering when Mr.C would come around to focusing his prescience in a column about the Singularity.

What do you mean, 'smarter than humans?'

Intelligence is a quantity, something measurable by I.Q. tests, while smartness is a quality (as is stupidity). And a quality is a human judgement which has no understandable meaning outside of the human race.

For example, suppose you try to rob a bank. If you get caught, the exercise was stupid. If you get away with it, maybe not so stupid. This will be true regardless of how much or how little intelligence you applied to robbing the bank. so intelligence will have little to do with whether robbing the bank was smart or stupid.

Furthermore, it's only smart or stupid in human terms: human beings think it's nice to have money without working for it and not so nice to have to go to prison. This is how we judge what is smart and what is stupid.

For another example, someone tells you to ask a certain person out and you marry that person and then live happily ever after. How could you have been told anything smarter? Conversely, if your marriage didn't turn out so well, how could they have told you anything stupider? Smart and dumb are purely judgements about quality and have no quantitative meaning.

So how can the 'singularity' make any kind of judgement? It would simply be pure qualitative intelligence and wouldn't have any feelings--so how could it care about anything? Why would it worry about going to prison? Why would it lust after money? Why would it make any difference to it if your marriage was happy or not?

If the singularity doesn't have any feelings, then nothing should motivate it, no matter how intelligent it might be.

If any singularity ever actually occured (which for more practical reasons I highly doubt) it would just be something that would be extremely intelligent in the sense that it would be able to answer any question we might put to it. But it doesn't follow that such a thing would necessarily be something that would need or want to do anything on its own. Hence the singularity ought to be completely safe to us.

woooo, hold on here. RobX ....did you read Kurzweils entire book? You left out the part about computers being grafted onto the human mind and making IT 10,000X more efficacious.

I understand your conservation of difficulties arguement. I just think in this particular case it serves to camoflauge the more important point. The Kurzweil singularity as called out in his book is about 2040 and is when the intelligence augmentation mentioned above kicks in to such a degree that progress will be absolutely dizzying. Scientific research formerly requiring years will occur in minutes.

I have no doubt at that point swindles such as Man made global warming will be clearing unmasked in the eyes of all and thus new swindles will have to be originated. This time for the sake of entertainment as our collective minds will watch the numbers on the Gore electric meter / bank account and see the fraud unravel as quickly as it is perpetrated.

This article seems to broadly miss the point of the singularity. Sentences like “…by people and machines who may not have yet grown to Singularity, but what will really change everything is when the price of Singularity drops low enough to apply to the computer on our desks or on our wrists,” demonstrate a woeful lack of comprehension of the topic at hand. Sorry to flame, but it’s true.

The Singularity isn’t like a 7.1 stereo system, something that you upgrade to. It is by definition the point beyond which reasonable prognostication is no longer possible. I think Cringely needs to re-read his Kurzweil. For someone who so often has something interesting and insightful to say, this column was a huge disappointment.

A nice little talk in Kurzweil's own words can be found here http://www.ted.com/index.php/talks/view/id/38.

I just read Charlie Stross' Accellerando, and have put a little subway thought into what this means. His (Stross') vision of the future is exciting and chaotic and full of change, but it's interesting in how it stands almost in opposition to what I think will actually happen.

Assuming computers remain mostly passive, and by extension benign, there's a good chance that humanity will be granted it's fondest wishes - infinite, high quality food on demand? Cheap space tourism? Maybe all that and more, but what I really think is going to happen is nothing. I think this because I beleive that humanity's fondest wish is that Everything Stays The Same. People are scared of change - they're scared of the future and scared of the unknown.

I think people want cars that get 1000 mpg, not instantaneous teleportation - just too different and scary and unknown. People want a million dollars - but not a radically new type of economy that may eliminate want.

People also really want to be able to come home and draw the shades and have privacy - something that's going to be virtually elminated in virtually all post-singularity scenarios.

I don't really have a point. Just wanted to get some of that out.

“Robert” -- from George Margolin you old Inventor Acquaintance.

Your comments and Kurzweil’s book brought me back to my early teens when I was pondering the nature of life and the universe (as well as the mysteries of women – which I understand as little now at age 77 – as then at 17 or earlier) – I was an avid SF reader. And “what’s next” was a common story plot.

But to me – having always been a techie – ergo my profession as an inventor – I had and have – always considered the inescapable PROBABILITY that WE (humans) are THE INEVITABLE EVOLUTIONARY STEP needed for the creation of the NEXT step in self-aware creatures. I don’t know what to call this next stage – It’s been labeled “Omega Man” and Man-Machine – but IT WILL HAPPEN whether man IS or is NOT part of that superseding society.

So I didn’t need the Brilliant Mr. Kurzweil to tell me that. Though he partly confirms my own long held expectation. But I’m not placing any moral correlation of coloration on it.

To quote the great scientist, Doris Day and her song “Che Sera, Sera – Whatever will be, will be” and the MARVELOUS WRITER William Day – from his ground breaking unknown book “Genesis on Planet Earth” --- self published by “The House of Talos” in 1979 and updated and RE-published by Yale University Press in 1984 (or so) – He also wrote brilliantly about the origin of life on earth AND – the EXTREME probability that we are the NECESSARY link to our successors.

I’d like to recommend this book to you and your readers. It can be found for PENNIES – on ABE Books and even on Amazon. But get the second edition from Yale.

Love your stuff.

George Margolin
patentor@gmail.com

It’s a bit sad that nobody seems to have remembered that Arthur C Clarke’s book “Profiles of the Future” first published way back in 1962 presented the development of the truly intellingent computer as ‘The last invention man need make’.

Its a stunning little book that looks at why people failed spectacularly to forsee the future and then in fun fashion, Clarke attempts to do so himself. He lists inventions and discoveries of such things as ‘immortality’ and ‘climate control’ in chronological order along with the dates at which he thinks they might occur.

Because it was first published in 1962, you can look at his list and see how his predictions have fared so far. His thoughts are certainly worthy of analysis in this discussion.

A short review of a revised version of “Profiles” can be found here

http://www.infinityplus.co.uk/nonfiction/profiles.htm

Zack, your comment:
"People want a million dollars - but not a radically new type of economy that may eliminate want."
is the wisest one I have seen here and shows the best understadning of human nature.

The acceleratiing return we see from technology is real. But the "singularity" I see is a collision between human behavior and the power of our increasingly rapid accumulation of knowledge.

The Fermi paradox comes to mind. It may be quite rare for any civilization to survive this stage of development. I do not suggest that technological progress should be impeded, but I can't see a long term future for us unless human behavior is radically altered in an altruistic direction. The time between now and that point holds extreme risk.

In the world of Letterman, is it something or is it nothing?

This is nothing.

Don't know where to start with this!

As a species, we've been discussing this topic for about a century, via SF, but ALL that is speculation, & could go in any of the directions, or none.

The BitTorrent scenario rings true, because of what has ACTUALLY happened - I'm think about the late Eighties & early Nineties, when PCs would come with compression software to make more space on their hard disks - eventually, disks got bigger & cheaper, faster than we could use the space.

I think, as usual, the development of our societies will lag behind the effects of innovation, unless, of course, we apply the post-singularity intelligences to social development, assuming they'll accept our requests!

Huxley's "Brave New World" springs to mind. And I wonder if, for the masses, the Singularity hasn't already happened. How many millions are chained to their TVs, will buy whatever TV tells them to buy, worry about whatever "news" TV tells them to worry about, and believe everything that the Great TV tells them to believe. I will go and live in a mud hut with the Savages somewhere "inside the Reservation" (which actually is "outside the Brave New World", where I might die from Malaria or some other curable disease, but at least I am free.

Based on your previous articles on the Telecom failures to provide what the American people have paid for, the only question will be: "How much for 'Singularity-Plus' and when can I have it installed?" This will just be a marketing problem. That is until a 14 year-old figures a way to strap multiple Singularity modules together to make something uncontrollable. Sounds like a "King for a Day" scenario to me.

Don't worry...if the supercomputers of 2029 get uppity, we can just install Microsoft Office 2024 on them and bring them to their knees.

so many assumptions... especially on the nature of intelligence. we keep thinking we know how it works, but at every step we see a new level of complexity. e.g. genes don't just code proteins, they code proteins that activate other genes etc. who knows how many other levels of complexity there are? it may in fact be beyond our ability to understand enough to replicate it.

we can't even have an office suite that doesn't crash. and we all know how much intelligence is in there...

I, too, am absolutely confident it will happen.

We will not realize it is happening when it does--like a cavity, or growing old, one day we will realize, individually and of our own accord, that our reality has dramatically shifted.

The joysticks will be shifting us.

".......Don't worry...if the supercomputers of 2029 get uppity, we can just install Microsoft Office 2024 on them and bring them to their knees......."

Best comment all year....

I doubt that something like the 'Singularity' will ever come. Rather, we'll more likely have numerous specialty devices - that, from an A.I. perspective, are AI's, but only in specialty areas. Perhaps some devices will have more than one of these AI's built into them - every day computers will likely have voice recognition AI's built into them eventually, which will likely replace some aspects of how we interact with computers (I think Star Trek gives us a good view on this), but ultimately, AIs will never achieve much, and never be self thinking outside of their original programming. (I.e. they won't ever evolve in their thoughts or do things they they were not programmed to do by humans.)

So, look for increased functionality in computer systems, but no other real advantage, and no new life forms.

Roger Penrose's book 'The Emperor's New Mind' is an interesting look at artificial intelligence-- winner of two Nobel's, he discusses the nature of intelligence and self-reflection and ties it to quantum phenomena. He strongly argues against artificial intelligence ever being successfully developed-- I think the machines will hit crippling developmental obstacles that will level off technological progress and limit any potential for self-awareness---

P2P is not just a way to cope with current limitations, it is really another way to use the network.

It enables N*N connections where client-server only enables N connections. So at any time, in a any future of the Internet, P2P will still be relevant.

Anyway, the suitcase is always full, and the hard drive is always full. So I'm sure we'll find some pretty heavy stuff to transfer digitally. Say a teleportation of your whole body's patterns - how much would that weight to make you feel good about it?

How many GB/s to become yourself again in less than 5 seconds?

Ummm, Roger Penrose has not won any Nobel Prizes yet.

"The Singularity may well bring with it the end of death, but I am 100 percent certain that taxes will survive."

Mr. Cringley, if I can obtain a device that is 10,000 times smarter than humans think of the tax avoidance schemes it can devise.

Of course the Government will try to make the Internal Revenue Singularity (IRS) much smarter than what is available on the street to devise a tax code without loopholes; however due to the lead and lag of systems and ingenious hackers most likely one would be able to buy the latest tax scheme, that while lasting only nanoseconds would last long enough to help get one's taxes to the "blessed number" -- 0.

Exponential growth cannot last forever. You can only go through 256 doublings, and then you're out of universe.

All curves are S-curves. They look exponential in the beginning, but then they level off.

It was easy for me to expect processor clock rates to go up by around a factor of six every 5 or 6 years: 1MHz in 1983; 8 MHz in 1988; 60 MHz in 1994, 400 MHz in 1999. Now that's stopped (except for the non-performing 3+ GHz Pentium IV), and the industry is going to multiple cores.

An entry-level PC used to always cost $1500, from a late-1970s Apple II to budget Compaqs. Starting a few years back, prices started to fall when consumers couldn't be persuaded to shell out for the bigger and better and faster and fancier machines. Now I see ads for Dells with 19" flat panel displays for $499. Yes, the speed of rendering 3D scenes is mediocre (compared to more expensive machiens; it blows away state-of-the-art machines from ten years ago...)

The market might not sustain continual increases in technological capabilities.

I don't think we'll see the singularity as described by Kurzweil and feared by Joy. What would those "super smart" machines do? Surf the web? Make up stories? How could what they write/script/direct/render be any good, make an emotional connection with the viewer?

Aside from information/entertainment, these super-intelligent machines would have to do something for people to bother to build them. A super intelligent PDA with super networking would do what? Enable a user to talk to other people (video conference, even) access, create, and share documents, email messages, audio and video recordings. You might even be able to ask it to compose new music in the style of Mozart. But poetry in the style of Whitman? Where would it get the understanding of emotion to do that?

the following is an excerpt from an essay called "the singularity is always near" by kevin kelly. most insightful thing i ever read on the subject:
"Fourth, and most important, I think that technological transitions represented by the singularity are completely imperceptible from WITHIN the transition that is represented (inaccurately) by a singularity. A phase shift from one level to the next level is only visible from the perch of the new level -- after arrival there. Compared to a neuron the mind is a singularity -- it is invisible and unimaginable to the lower parts. But from the viewpoint of a neuron the movement from a few neurons to many neurons to alert mind will appear to be a slow continuous smooth journey of gathering neurons. There is no sense of disruption, of Rapture. The discontinuity can only be seen in retrospect.

Language is a singularity of sorts, as was writing. But the path to both of these was continuous and imperceptible to the acquirers. I am reminded of a great story a friend tells of some cavemen sitting around the campfire 100,000 years ago, chewing on the last bits of meat, chatting in guttural sounds. One of them says,

"Hey, you guys, we are TALKING!
"What do you mean TALKING? Are you finished that bone?
"I mean we are SPEAKING to each other! Using WORDS. Don't you get it?
“You've been drinking that grape stuff again, haven't you."
“See we are doing it right now!”
“What?”

As the next level of organization kicks in, the current level is incapable of perceiving the new level, because that perception must take place at the new level. From within our emerging global cultural, the coming phase shift to another level is real, but it will be imperceptible to us during the transition. Sure, things will speed up, but that will only hide the real change, which is a change in the rules of the game. Therefore we can expect in the next hundred years that life will appear to be ordinary and not discontinuous, certainly not cataclysmic, all the while something new gathers, until slowly we recognize that we have acquired the tools to perceive that new tools are present - and have been for a while." - an excerpt from an essay by Kevin Kelly

Owned_By_2_Cats,

you really should read Kurzweil's book(s). He addresses all of the points that you made. I'm not saying that I believe he's correct in all of his predictions, or even in the Singularity. Nevertheless he makes a well thought out case, and as outlandish as his ideas are, I've never read a compelling counter-argument.

Need more evidence? Try a Google search for "Blue Brain", and compare what is going on right now with Kurzweil's predictions.

The singularity has already occurred. Who can get by without the brain-boost given by Google anymore. The .boom to .bust cycle is shortening too. Machines are already better than humans at many many tasks.

i think you just jumped the shark

People confuse "Artificial Intelligence" with "Artificial Human Intelligence". We imagine an AI resenting doing our bidding because that's how we'd feel if we were in it's place.

I didn't think much about the Singularity until recently. The endless disappointing dead ends of AI research made me think it must be very far away.

Jeff Hawkins' "On Intelligence" changed my mind. If he's right then AI research has been using the wrong approach. Either way we'll know soon enough because he used the money he made inventing the Treo to fund first the Redwood Neuroscience Institute and then to found a new company called Numenta to develop intelligent machines based on his theory of how the brain works.

I've already downloaded his open source AI software! It's pretty cool.

You want to know what we'll do with our super-smart post-singularity technology? You need look no further than the Middle East, Iraq, Darfur, or Zimbabwe to see what we're doing with whatever pre-singularity technology comes to hand.

I agree that we will have serious trouble adjusting to a post-singularity world, but I see no reason to assume Moore's Law will survive.

In fact, isn't that the whole reason is is called a singularity? Once you get your hardware to improve itself, the growth curve should become steeper and steeper each generation. And, yes, this will be limited to supercomputer centers at first, but as soon as that supercomputer has designed the next generation, there is no reason not focus on cost optimization while we slow humans struggle to build that new computer.

Of course, we could also just give it a robot arm, a bucket of white sand and some duct tape and tell it to get on with it...

the following is an excerpt from an essay called "the singularity is always near" by kevin kelly. most insightful thing i ever read on the subject:
"Fourth, and most important, I think that technological transitions represented by the singularity are completely imperceptible from WITHIN the transition that is represented (inaccurately) by a singularity. A phase shift from one level to the next level is only visible from the perch of the new level -- after arrival there. Compared to a neuron the mind is a singularity -- it is invisible and unimaginable to the lower parts. But from the viewpoint of a neuron the movement from a few neurons to many neurons to alert mind will appear to be a slow continuous smooth journey of gathering neurons. There is no sense of disruption, of Rapture. The discontinuity can only be seen in retrospect.

Language is a singularity of sorts, as was writing. But the path to both of these was continuous and imperceptible to the acquirers. I am reminded of a great story a friend tells of some cavemen sitting around the campfire 100,000 years ago, chewing on the last bits of meat, chatting in guttural sounds. One of them says,

"Hey, you guys, we are TALKING!
"What do you mean TALKING? Are you finished that bone?
"I mean we are SPEAKING to each other! Using WORDS. Don't you get it?
“You've been drinking that grape stuff again, haven't you."
“See we are doing it right now!”
“What?”

As the next level of organization kicks in, the current level is incapable of perceiving the new level, because that perception must take place at the new level. From within our emerging global cultural, the coming phase shift to another level is real, but it will be imperceptible to us during the transition. Sure, things will speed up, but that will only hide the real change, which is a change in the rules of the game. Therefore we can expect in the next hundred years that life will appear to be ordinary and not discontinuous, certainly not cataclysmic, all the while something new gathers, until slowly we recognize that we have acquired the tools to perceive that new tools are present - and have been for a while." - an excerpt from an essay by Kevin Kelly

The following is an excerpt from an essay called "The Singularity Is Always Near" by Kevin kelly. Most insightful thing i ever read on the subject. The remainder of this post is a direct paste from one of his blogs, called The Technium:

Language is a singularity of sorts, as was writing. But the path to both of these was continuous and imperceptible to the acquirers. I am reminded of a great story a friend tells of some cavemen sitting around the campfire 100,000 years ago, chewing on the last bits of meat, chatting in guttural sounds. One of them says,

"Hey, you guys, we are TALKING!
"What do you mean TALKING? Are you finished that bone?
"I mean we are SPEAKING to each other! Using WORDS. Don't you get it?
“You've been drinking that grape stuff again, haven't you."
“See we are doing it right now!”
“What?”

As the next level of organization kicks in, the current level is incapable of perceiving the new level, because that perception must take place at the new level. From within our emerging global cultural, the coming phase shift to another level is real, but it will be imperceptible to us during the transition. Sure, things will speed up, but that will only hide the real change, which is a change in the rules of the game. Therefore we can expect in the next hundred years that life will appear to be ordinary and not discontinuous, certainly not cataclysmic, all the while something new gathers, until slowly we recognize that we have acquired the tools to perceive that new tools are present - and have been for a while.

I think one big thing that readers and commenters have not considered much is the coming singularity in genetics and gene manipulation. As the Human Genome Project illustrates -- and discussed in the original essay -- genetics (and coming genetic changes to the human genome) are tightly coupled with computers. I don't know if computers will ever be "sentient" (what ever that really is...) but computers will definitely be the tool by which the human race achieves singularity (again, what ever that really is...). When we can truly take charge of our own genetic destiny (and the rest of life on this planet), all bets are off, and the future is beyond prediction. I like the comparison to an event horizon: There's a singularity out there, but because of it's very nature, there is a horizon beyond which we cannot now see. Read Hawking for how singularities are perceived, both by observer and participant.

The erros and misstatements posted here are so legion that addressing each one would be futile. Please, no more posting of the "Singularity is Always Near" comments. We saw them the 1st, 2nd, 3rd...time. The entire article (and response) can be found (one time only) on Kurzweil's site.

The fantasy aspects vanish as advances become part of our daily lives. Talking & seeing anyone, anytime, anywhere is now ho-hum. The increasing rate of discovery in every area (matter, energy, biology, AI) is simply stunning. A good place to begin - and one grounded in reality - is the site for the Singularity Institute. Hours of talks by scientists, inventors, business leaders, philosophers, believers and doubters are present. The tone is civilized & mostly with little hype. We should at least consider the possibility if only because those who predict its arrival have been so prescient in the past.

Unless one believes consciousness derives from non-material sources, i.e. God, there is no logical reason why matter other than our brain cannot one day attain self-awareness thru means other than evolution. After all, cave woman did not wake up one day suddenly "aware". It was a long, multi-step process that had as much to do with language and culture as biology.

I just saw a very interesting Wed-Series about the Singularity...it seems like a GREAT t.v. show...check
it out on www.doplereffecttv.com

What about the "First Contact" aspects of the singularity? If machines achieve human-like or higher intelligence will they also achieve sentience (as in a sense of self)? What would the consequences to our society be when the masses realize we have built a form of consciousness with its own self interests. And if these entities interact, how will their society form and interact with ours? The transition period will evolve into a decades-long process of ethical and moral resolutions, diplomacy and negotiation. If a computer system becomes sentient, is it ethical to cut the power? What does the system owe the power company in that case? How will sentient computers be compensated for their labor?

The economic and social turmoil that will be triggered by the singularity sounds way too much like "interesting times" for my benefit. I doubt the turmoil will end in less than a century after the singularity is reached, unless an early decision is made to prohibit attempts to achieve it.

My predection is that long before we reach "the Singularity" we will invent something akin to the Star Trek "holodeck", at which point all meaningful progress will cease (as every member of the human race will be inside the holodeck, indulging their favorite fantasy)

When will AI surpass human-level intelligence?

[ ] 2010-20
[ ] 2020-30
[ ] 2030-50
[ ] 2050-70
[ ] 2070-2100
[ ] Beyond 2100
[ ] Prefer not to make predictions
[ ] Other: __

Results:
http://www.novamente.net/bruce/index.php/?p=54

I just skimmed some of the comments and I'm getting that alot of people want to poo-poo the singularity. The only problem is WE are biological computers. Thus the singularity has already been proven and that whatever we can do, others can do whether it be by evolution or revolution. Thus the well thought out hit on the singularity is just the output of a THINKER personality type. However I as a SENSOR do not get lost in forest because of opaque trees. I can clearly see that processing power will lead to intelligence just because I don't over think it. I go straight to the point. We exist and thus the described effects of the singularity will exist. Intelligence will lead to personality and a hop skip jump to existance.

Worrying about Singularity is like worrying about your untied shoelaces as a freight train is bearing down on you. We can be nicely extinct all by ourselves without no automated intervention.

Dear Mr. Cringely and cpmmentators,

Very interesting discussion.

Here are my thoughts about Singularity
if you care to read:

http://www.ghandchi.com/479-KurzweilFuturismEng.htm

All the Best,
Sam Ghandchi
http://www.ghandchi.com

Bah, this is complete drivel. DARPA is hoping that, by 2100, computers can equal the intelligence of a honeybee. Even given an accelerating rate of progress, which is unlikely, computers won't equal the intelligence of a 4-year-old child for another two or three thousand years.

Artificial intelligence has been on the books for a long time. Problem, memory, speed, programming. You can with brute force solve a lot of problems, parallel processing(good for long branching chess, routing problems, matrix inversion). But the subtleties of language, thinking(actually), not mechanical branching, voice, body language, all seems to still be missing in any program or other system. I wouldn't hold my breath(Japanese robots - so). Talking computers we'll just turn 'em off after the novelty wears off. It will be the same with any computer mechanical life form, tedium will take over and it will still be a lot more interesting to communicate with living life forms. Why maybe it's embedded in our DNA. After 35 years of working with computers, give me a good old card game, opera or theatre with real people.

Technological improvement does not map to intelligence. The brightest minds in AI still can't tell you what Artificial Intelligence is. This is because they can't tell you what intelligence is. We just don't know if we need faster/bigger computers to create intelligence. We need to have an epiphany. Those tend not to follow time lines.

Pat O

The Internet is just a fad.

I used to think that THE NET was one of the 3 greatest inventions of the world, somewhere in between electricity and the internal compustion engine. But, I was wrong. THE NET has fallen to TV standards, and without a NET, computers and Singularity - are just more junk.

We can't even trust them to VOTE.

May I suggest a TAX REVOLT to prevent THE MAN from imposing Singularity on us all?

The Internet is just a fad.

I used to think that THE NET was one of the 3 greatest inventions of the world, somewhere in between electricity and the internal combustion engine. But, I was wrong. THE NET has fallen to TV standards, and without a NET, computers and Singularity - are just more junk.

We can't even trust them to VOTE.

May I suggest a TAX REVOLT to prevent THE MAN from imposing Singularity on us all?

"For the backbones that time looks to be around 2013 and for our home connections to catch up will take until 2016".

It will not matter. The alien settlers will be back on earth on December 2012. Fox M