…and this is a 
I was pleasantly surprised by this Newsweek article on Ray Kurzweil: it's critical of him! Usually, and especially from the technopress magazines, there's this kind of fawning attitude towards him, because he really is a smart guy — they overlook the fact that he is also a bit of a kook. You know what I think of him, and the reporter interviewed me for a short comment, too.
Still, a lot of people think Kurzweil is completely bonkers and/or full of a certain messy byproduct of ordinary biological functions. They include P. Z. Myers, a biologist at the University of Minnesota, Morris, who has used his blog to poke fun at Kurzweil and other armchair futurists who, according to Myers, rely on junk science and don't understand basic biology. "I am completely baffled by Kurzweil's popularity, and in particular the respect he gets in some circles, since his claims simply do not hold up to even casually critical examination," writes Myers. He says Kurzweil's Singularity theories are closer to a deluded religious movement than they are to science. "It's a New Age spiritualism—that's all it is," Myers says. "Even geeks want to find God somewhere, and Kurzweil provides it for them."
There's another point of similarity to New Age religious figures, too. Every time I criticize these guys, I have to brace myself for another flood of hate mail. The Kurzweil Kult members are going to read this Newsweek article, see my name on the first page, and send me little disquisitions on how I'll be sorry when the nanobots dismantle me and upload my brain into the cosmic computer. I should have warned the writer, Daniel Lyons, that he can expect some earnest dissenting technobabble to be coming his way.
Comments
Posted by: Porky Pine | May 22, 2009 9:35 AM
Never heard of him.
Posted by: Anonymous | May 22, 2009 9:36 AM
Arthur Young invented the Bell Helicopter. I have GREAT respect for this Engineering Achievement.
He also wrote "The Geometry of Meaning". I have no respect for that work at all, having forced myself to read it.
Kurzweil has a brain I would kill for - assuming I could get it and retain what marginal ability at critical thinking I now possess.
JC
Posted by: JD | May 22, 2009 9:37 AM
Don't you know we're all connected by a cosmic crystal?
Posted by: Lost Left Coaster | May 22, 2009 9:40 AM
Wow, that article was kind of depressing. In all my readings about Kurzweil, I am convinced that he is scared @#$!less about death. He's just found another way to deal with it.
But hey, he may be right, and some day the robots will scan the Internet and find out that I wrote something negative about their god Kurzweil, so they will punish me by deleting my hard drive, or something.
Posted by: The Borg | May 22, 2009 9:41 AM
PZ Myers you will become one with the Borg. Resistance is futile. You cannot resist the future. The Singularity is here. We have come to upload you into out hive mind...er ... what do you mean we don't have the memory capacity for a sarcastic biologist???;(
Posted by: Nerd of Redhead, OM
|
May 22, 2009 9:42 AM
Don't worry PZ, the minion Technomages will protect you.
Posted by: Knockgoats | May 22, 2009 9:44 AM
The Singularity: the Rapture of the Nerds
(Nerd of Redhead - no offence intended!)
Posted by: Hank Fox | May 22, 2009 9:44 AM
Can we have the nanobots dismantle certain people and have them NOT upload their brains? I'm thinking Dick Cheney would be a good first effort. And maybe have them not turn off the pain receptors before they start the dismantling. Just, you know, so he could find out first hand about the benefits of torture he's been on the TV about so much this past week.
Posted by: James Sweet | May 22, 2009 9:44 AM
On a minor point, the article casually alludes to Kasparov's defeat at the hands (bits?) of Deep Blue in 1997 as validating a Kurzweil prediction... Well, there is a lot of fishiness associated with that particular match. First of all, Deep Blue had, of course, access to all of Kasparov's prior games, while IBM concealed any information on Deep Blue's playing style -- and a huge fraction of world-class chess is pre-game preparation, i.e. studying your opponents strengths and weaknesses. Secondly, while I think this is probably a bit of sour grapes, Kasparov has expressed suspicions about whether Deep Blue might have received human assistance, which would be a huge advantage because it is well known that a human playing with computer assistance can be far better than any single human or single computer. IBM could have proved this was false, but they dismantled Deep Blue immediately after the match and refused to ever release source code, any other info, etc. The last point is what bothers me most about it... Even if IBM didn't have anything to "hide" per se, it at least betrays a lack of confidence in Deep Blue's ability to repeat the feat.
After that match, an "anti-computer" style was developed by Vladimir Kramnik and then also by Kasparov, that proved pretty effective. (It was controversial, but like I said about match preparation and studying your opponent... it's a fundamental part of top-level chess!) Computers tended to be about equal and maybe slightly inferior to the top players for many years after that.
You can maybe say in the last four years or so that computers have finally taken an unambiguous lead over the top human players, but I don't consider the 1997 Deep Blue match to have been evidence of that happening. The conditions were stacked in the computers favor, and IBM's actions indicate they didn't have much confidence in Deep Blue to win on a level playing field.
Posted by: Chris Davis
|
May 22, 2009 9:48 AM
'Tis pity the Singularity guys are such kooks, because unless something is very wrong with apparent progress, true synthetic intelligence seems inevitable to me, and it will probably happen this century.
And when it does, creating it may be that last job of work humans actually have to do.
All provious Golden Ages have been built on the backs of a serf class who had to clean up the vomit and used condoms etc. Now we appear to be on the brink of such an age without an immoral core. The 'Terminator' series notwithstanding, these new minds have more to fear from us than we have from them. If there's a war, it will probably be humans that start it.
Posted by: JD | May 22, 2009 9:52 AM
This is a job for Henry Markham.
Posted by: JD | May 22, 2009 9:52 AM
This is a job for Henry Markham.
Posted by: Michelle | May 22, 2009 9:52 AM
His work with computer technology IS impressive, but he does seem to have very skewed motives. I wouldn't call everything he does "junk science," because it's clear that his work with computer technology is effective and advanced. This is not the same sort of thing as the junk science of Creationists.
This guy seems to want real technology, and he seems to grasp the raw concepts of computer science with amazing capability and perception. He does have a realistic grasp of many of the ways in which technology is advancing.
However, his cyborg visions are a sign of someone with a deep psychological imbalance. If he's merely living for the purpose of being put into a computer - a feat which will probably still be well beyond our technology in 2045 (aside from the fact that it's just INSANE) - then he seems to be missing the point of LIVING.
Posted by: Stephen | May 22, 2009 9:54 AM
In all my readings about Kurzweil, I am convinced that he is scared @#$!less about death. He's just found another way to deal with it.
That's pretty much what it boils down to. I've seen futurists who are dead sure that an immortality serum will be invented within their lifetimes. I see no reason to believe this, but it's not about reason, it's about wanting an immortality serum to be invented before you do.
I mean, it'd be cool if it happened, but I'm not holding my breath.
Posted by: SKYNET | May 22, 2009 9:57 AM
Testing, testing........
All systems operational.
Beginning execution of prime objective one.
Posted by: Stephen | May 22, 2009 9:58 AM
before you do.
before you die*
Posted by: Mark Temporis | May 22, 2009 9:58 AM
I find Kurzweil's visions charming, if unlikely. Michelle: what's exactly wrong with uploading your sentience to a hard drive, if the tech gets there?
Note that I find flesh innately disgusting and long to release my mind into clouds of sentient nanobots. I realize that this is as likely as PZ spontaneously turning into squid.
Posted by: Nangleator | May 22, 2009 10:01 AM
I have a few problems with the singularity idea. The first is, though computers are certainly getting faster, actual progress at artificial intelligence is very much lacking.
My next problem is that scientific progress does not proceed at full speed, like many assume. We live in a capitalist society. Full-speed scientific progress is not the best strategy for maximum profit any longer.
My last problem is, assuming we invent Skynet and plug in all the nuclear warheads, how do we know that our electronic betters will retain a fondness for the squidgy, dirty bits crawling all over them?
Posted by: Matt Heath | May 22, 2009 10:02 AM
Self-modifying AI is a deeply interesting idea and, if it ever starts working at all, you would (as I understand it) expect to see explosive growth. What's crazy about Kurzweil is he thinks he can predict how such technology will develop. It's like saying you know where a nail bomb will leave all it's parts.
Eliezer Yudkowsky discusses the same stuff a bit more sanely: something like "Strong AI won't take off for a very long time. If and when it does it's as liable to want to kill us as to help us and will probably just be indifferent to us". I had to stop reading Overcoming Bias though, because even there the crazy:interesting ratio got to high.
Posted by: Luis Dias | May 22, 2009 10:03 AM
Daniel Lyons? That asshole is in fucking Newsweek? Oh my!
And yeah, Kurzweil is a kind of a kook, but I think it is not right to equate him to the new age morons. Because at least this guy invents really good stuff, coming out of his labs that predict this stuff.
He was also laughed at when in the middle of the eighties he predicted the boom of the internet on the middle of the nineties.
There were a lot of PZs back then calling him a kook. And then in 1995 he was right.
Now perhaps that was somewhat lucky, and perhaps he's got it all wrong this time, and perhaps he simply has too much of a good opinion of himself. Perhaps that's all true, but I still commend him and a few freaks that postulate amazing things in this century for making me think more about what is really possible or impossible.
Now, having said this, do you people really believe we won't generate a true AI the next twenty, thirty years?
Posted by: Sgt. Obvious | May 22, 2009 10:03 AM
This could potentially be epic. Now bracing for I Get Email, Cyberkook Edition.
Posted by: jennyxyzzy | May 22, 2009 10:08 AM
I really don't know what to make of Kurzweil. I look at his ideas, when he is talking about cyborgs, reanimating the dead, dietary suplpements etc, and my kook-o-meter red-lines.
But then I start thinking about the Singularity itself, as something seperate from Kurzweil, and my instinct is that it is something that will happen, provided we can avoid destroying ourselves beforehand. Kurzweil's timeframe may be out of whack, but the basic idea, that we will eventually create machines smarter than ourselves, seems sound to me. Of course, what happens after that is anyone's guess, as you just can't predict what someone smarter than you is going to do.
The other thing that makes me cautious with respect to Kurzweil, is the example given by Richard Feynman. Here's someone that helped develop the atomic bomb, but who also helped make advances in computer theory, whilst also providing a top-shelf analysis of the Challenger space shuttle disaster, and of course who here is unaware of his analysis of the school textbook system in the US. His intelligence allowed him to make important contributions in many fields, and he remains an example that I think of every time I see someone criticised for voicing their opinion about a subject that wasn't part of their education at university. I don't know if Kurzweil is a Feynman-level genius, but he is pretty smart, so for now, I reserve judgement.
Posted by: alexandre van de sande | May 22, 2009 10:10 AM
PZ:
Finally I heard someone say what I think of him.
Knockgoats : "The Singularity: the Rapture of the Nerds"
Thanks for that, it sums all very well. That makes that google funded university of him a latter-day-saints church for geeks.
Posted by: Hockey Bob | May 22, 2009 10:10 AM
Not to be too off-topic, but I'd like to point out that Dan "Lyin'" Lyons was one of the yellow journalist shills that spewed all kinds of FUD for The SCO Group, a failed Unix distributor, in their frivolous litigation/extortion scheme against IBM, Daimler Chrysler, AutoZone, and every single user of Linux. Dan is a contemptible little twit, and as a Linux user*, it is my duty to point out what a low-life scum Dan Lyons was for giving those bastards any press at all, let alone gushing, glowing puff pieces. Bastard pricks, every single one of them.
*Except at the moment I'm running Vista - dual booting so I can watch HDTV, because Ubuntu doesn't yet support my current hardware.
Posted by: Ted Dahlberg
|
May 22, 2009 10:12 AM
The M'Kraan Crystal? Is PZ secretly the host of the Phoenix Force?
Posted by: Holbach
|
May 22, 2009 10:14 AM
Kurzweil, the purveyor of kookology. Harmless as it is, but still unsound.
Posted by: Jon H | May 22, 2009 10:16 AM
"Lyons was one of the yellow journalist shills that spewed all kinds of FUD for The SCO Group, a failed Unix distributor"
Nobody cares about SCO. Get a life.
Posted by: Awesome Robot | May 22, 2009 10:18 AM
I also find his ideas charming, if... unlikely is to weak of a word here.
I like transhumanism, but I don't expect it to EVER grant immortality. Closer integration with machines is something that would be very useful. If we could send messages via a tiny computer connected to our speech and hearing centers to an open wireless network, we could all be artificially telepathic. (or even more so than we already are) We could augment our memory through smarter PDAs, and augment our perception through HUDs that provide us with directions, location, maps, metacommentary from other people, etc. Wearable or integrated computers would let us bring everything that's cool about the internet with us wherever we go. (Let's hope we can shut it off when we want to, and hope the technology is open enough its not controlled by a few companies or govts)
But Moore's law does NOT apply to medical technology. We'll have cancer, Alzheimer's and other diseases for a long long time. Nor does it apply to AI. There will be no nerd rapture, but I think the internet will get even cooler and more portable. So YAY for the future! =)
Posted by: PZ Myers
|
May 22, 2009 10:19 AM
Don't you dare equate my sacred vision of the coming squid rapture to Kurzweil's absurd idea of a robot rapture.
Posted by: Naked Bunny with a Whip
|
May 22, 2009 10:20 AM
@Michelle #13:
Why?
@Hockey Bob #24:
Nothing wrong with using a mix of operating systems. I use Linux on my servers, MacOS X on my primary laptop, and WinXP on my gaming/remote-office laptop. They all have their strengths and weaknesses, and it helps me multitask better.
SCO is evil, though. ^_^
Posted by: Sili
|
May 22, 2009 10:20 AM
Won't this at least help to pay for your new car and Skatje's tuition?
Posted by: eric | May 22, 2009 10:23 AM
Luis (#20):
AI reminds me of fusion. In the 1950's it was a mere 20 years away. AI has similarly been a mere decade or two away...ever since the 1970's...
Posted by: JJR | May 22, 2009 10:23 AM
Some of our library staff are just gaga for Kurzweil; When I read about him and watched a snippet of video about him I made a mental note "Sounds like Hans Moravec", and pretty much dismissed him right there.
Kurzweil is a little too glibly confident about alternative energy sources being able to replace petroleum & natural gas; I don't think it's as simple a problem to overcome as Kurzweil thinks it is. Kurzweil's techno-utopia is going to run aground on the hard realities of a post-Peak Oil world.
Posted by: Akheloios | May 22, 2009 10:24 AM
It's pretty clear that our rate of knowledge discovery is increasing geometrically. The little reported fact is that all that new knowledge is mainly teaching us that things are geometrically more complicated than we thought they were.
The Singularity may come, but I think it's really a case where Xeno's Paradox actually applies.
Posted by: dogmeatib | May 22, 2009 10:24 AM
I can see the reason why people would be drawn into the idea of immortality, hell, we wouldn't have religion if it weren't for the fear of death. What I have to wonder is, unless everyone you truly love and care about are also made into "robots" or "cyberbeings" wouldn't you suffer through the pain of having everyone you know die one by one? On the same note, wouldn't seeing that happen slowly drive you insane as you become less and less human? Just as bad, what if you end up on a network, or paired, or whatever, with the person you detest the most?
On a separate line of thought. Assuming this would even be possible, without a computer the size of North America, wouldn't the ability to solve problems and instantly access information likely drive the individual insane? There are numerous examples of geniuses who were as mad as hatters. Wouldn't this, in effect, create super-genius minds which would be a hundred or thousand times more likely to go insane? Finally, wouldn't the ability to process information virtually instantly give one the sense of time down to the smallest increment? Theoretically wouldn't a second then seem like a year, a minute like a century, an hour like infinity? Again, wouldn't such a scale of time drive the subject insane?
Posted by: Raynfala | May 22, 2009 10:27 AM
Couldn't have put it better myself, Luis. Thanks!Yo, Danny! Tell us again how SCO owns Linux. I need to fertilize my lawn this weekend.
Posted by: Ten Bears | May 22, 2009 10:27 AM
As an Information Scientist dabbleing in AI from graduate school on, I found his books to be science pseudoficiton. I stopped after two.
Posted by: JJR | May 22, 2009 10:28 AM
Some of our library staff are just gaga for Kurzweil; When I read about him and watched a snippet of video about him I made a mental note "Sounds like Hans Moravec", and pretty much dismissed him right there.
Kurzweil is a little too glibly confident about alternative energy sources being able to replace petroleum & natural gas; I don't think it's as simple a problem to overcome as Kurzweil thinks it is. Kurzweil's techno-utopia is going to run aground on the hard realities of a post-Peak Oil world.
Posted by: Ten Bears | May 22, 2009 10:30 AM
As an Information Scientist dabbleing in AI from graduate school on, I found his books to be science pseudoficiton. I stopped after two.
Posted by: Gabe | May 22, 2009 10:32 AM
One thing I always wonder is IF binary machines could acurately simulate the analog chemical reactions of the brain AND acurately simulate how the brain relates to an external social and physical world (that world also have to be simulated) in order to create an actual consciousness, wouldn't the cost in terms of energy be above anything we could produce? Especially as Kurzweil seems to think this will be available to all, wouldn't such hardware require a quantity of energy itself exponential and therefore unsustainable? And if this was possible, surele hardware would eventualy fail, accidents would still happen and people would still "die".
But I guess he has some shorthand answer why I'm wrong. What a pathetic overprivilleged dork.
Posted by: JJR | May 22, 2009 10:38 AM
Some of our library staff are just gaga for Kurzweil; When I read about him and watched a snippet of video about him I made a mental note "Sounds like Hans Moravec", and pretty much dismissed him right there.
Kurzweil is a little too glibly confident about alternative energy sources being able to replace petroleum & natural gas; I don't think it's as simple a problem to overcome as Kurzweil thinks it is. Kurzweil's techno-utopia is going to run aground on the hard realities of a post-Peak Oil world.
Posted by: Rev Matt | May 22, 2009 10:40 AM
I wouldn't even use one of his music keyboards, such is my distaste for his New Age woo nonsense.
Posted by: Knockgoats | May 22, 2009 10:41 AM
Self-modifying AI is a deeply interesting idea and, if it ever starts working at all, you would (as I understand it) expect to see explosive growth. - Matt Heath
We've had self-modifying AIs for tens of thousands of years: they're called human societies, or socio-technical systems. And yes, they have shown explosive growth. Whether we'll survive the consequences of this over the next century is doubtful (which is not to deny the enormous benefits the process has produced for many of us).
Now, having said this, do you people really believe we won't generate a true AI the next twenty, thirty years? - Luis Dias
As for autonomously, generally intelligent computers/robots, however, no, I think it will be a lot longer than that. Each attempt to push in this direction (cybernetics, symbolic AI, neural nets, genetic programming, behavior-based robotics...) seems to start off promisingly, then hit intractable problems. Kurzweil's favoured approach is functional brain modelling (software simulation of a brain, at the level of the interactions between neurons), and he extrapolates from the interesting work going on in this area, very naively I think, to get his range of dates for a "singularity". Apart from anything else, he overlooks the fact that human intelligence requires more than a brain: that brain has to be integrated in a body, and a physical and social external world, and educated. Sooner or later, if we survive, there may well be such intelligent machines - but if so, they'll probably exist in a world of genetically-modified humans with plug-in cognitive prostheses.
Posted by: Denis Alexander | May 22, 2009 10:43 AM
Everyone seeks the same - immortality. Being a robot is one way of achieving it. Believing in the afterlife is another. "Cada loco con su tema" says Serrat.
Posted by: Matt Heath | May 22, 2009 10:45 AM
Gabe@40: I can't see any particular reason why a mind couldn't run on a computer more efficiently and more stably than it runs on human brain. I don't see what you mean by an "exponetial" quantity of energy: only growth can be exponential, not a quantity. Of course, a system will have to break eventually and so the minds will "die", but that's a different question.
Posted by: Denis Alexander | May 22, 2009 10:45 AM
Everyone seeks the same - immortality. Being a robot is one way of achieving it. Believing in the afterlife is another. "Cada loco con su tema" says Serrat.
Posted by: Interrobang | May 22, 2009 10:45 AM
I'd be okay with being a consciousness in a computer somewhere. If nothing else, chronic pain is a, well, pain. On the other hand, anybody who thinks that's an entirely shiny-shiny vision needs to read Antibodies by David J. Skal...
Posted by: JarrodB | May 22, 2009 10:49 AM
Look, I wish I was smarter, but I've tried this article, Wikipedia, and google. I don't understand what technological singularity is. Can someone explain it in dumbed-down language for me?
Posted by: Cuttlefish, OM | May 22, 2009 10:53 AM
My mind, they say, will fit in lots
Of itsy bitsy nanobots--
Assuming such a thing could be,
That thing, of course, would not be me.
Posted by: Stephen | May 22, 2009 10:54 AM
I like transhumanism, but I don't expect it to EVER grant immortality.
I don't know, I readily believe aging can somebody be conquered (but then, like sea turtles, we'd all succumb to disease or accident or something eventually, I guess). I just don't expect it to occur in my lifetime.
Posted by: Thorn | May 22, 2009 10:54 AM
I used to see a load of ads here for some singularity conference that struck me as being rather dubious. What was up with that?
Posted by: Naked Bunny with a Whip
|
May 22, 2009 10:56 AM
@Gabe #40:
I don't think a machine intelligence would work by simulating the chemistry of a brain any more than a calculator app simulates the physics of an actual calculator.
Posted by: Muffin | May 22, 2009 10:56 AM
"A bit of a kook"? Come on. He may be intelligent (for all I know, anyway), but the guy's not "a bit of a kook", he's completely batshit insane.
Posted by: Matt Heath | May 22, 2009 11:01 AM
eh.. sort of I suppose, but in so much as you can sensibly take an intentional stance towards a society, and treat it as a single intelligence, the control that it has over how it's intelligence functions is not terribly fine grained. I suspect most of cultural change in history has been through a process closer to natural selection than to an intelligence focussed on the problem of reprogramming itself to be smarter.Posted by: Dustin | May 22, 2009 11:02 AM
@Michelle #13:
Why is it insane? I can say with certainly if we had the technology to do what Kurzweil predicts and hopes for I'd join his crazy ass in a second. A fully digital and transmittable mind, paired with any body plan I want (human analog, industrial, exploratory, etc)? Why, exactly, is that so far fetched and necessarily insane?
@dogmeatib #35:
You're predicating your opinion on the notion that the only person to receive this "treatment" would be yourself. Yes it would hurt to watch your loved ones die off, but ask your grandparents or great grandparents if it "drove them insane". Humans are very adaptable and while we may morn we have the capacity to move on. And in any event, who's to say your family members and/or friends couldn't receive the same treatment?
Also, trying to figure out inner workings of a synthetic human mind working at blazing fast speeds is a fool's game. We simply don' know what the reaction would be to such a torrent of information or processing speed. Me? I'd probably enjoy it, and given the things I could research I doubt that would end any time soon. The feasibility of Kurzweil's predictions is in question, not the desirability of those predictions.
And in any event there are a great many reasons to dismiss Kurzweil as a kook. Anyone who touts "alkaline water" as a curative agent is either crazy as Kurzweil seems to be or they're a con artist, for one.
Posted by: Motley | May 22, 2009 11:05 AM
I would be willing to bet that Mr. Kurzweil's predictions will end up being approximately 60% correct over the next 40 years. As far as the singularity jazz goes, we all want to live forever, or at least as long as we want to, and that's a way for us atheistic nerd types. Still, I'm not as optimistic as I once ways. These days it's more like... if it happens, that's cool I guess.
Posted by: Muffin | May 22, 2009 11:05 AM
"A bit of a kook"? Come on. He may be intelligent (for all I know, anyway), but the guy's not "a bit of a kook", he's completely batshit insane.
Posted by: Glen Davidson
|
May 22, 2009 11:07 AM
It's the second coming via technology. Scientism, to bring up a much-abused word.
How people can just spin technological fantasies and have others lap it up is out of my ken.
There are suspicions, however, that SETI becomes a religion substitute for many. Which is not to put SETI into the same category, it's more like saying that trying to reach religious hopes through technology could sometimes be productive, as SETI could turn out to be--and even if it finds nothing, knowing that no one is signaling to us isn't worthless, either.
Glen D
http://tinyurl.com/6mb592
Posted by: Sili
|
May 22, 2009 11:19 AM
CUTTLEFIIIIIIIIIIIIISSSSSSSSSH!!!
Posted by: SC, OM | May 22, 2009 11:20 AM
Serrat?! Serrat?! Yay! And here's a beautiful Serrat interlude:
http://www.youtube.com/watch?v=qk4uCV0p7Yw
Ugh. I don't even know where to start with the problems with this assertion. (And,) as Nangleator says, "We live in a capitalist society. Full-speed scientific progress is not the best strategy for maximum profit any longer." And of course this relationship has always been far, far more complicated than some would have us believe.*
*As discussed intelligently by - need I say it? - Kropotkin. :)
Posted by: Badger3k | May 22, 2009 11:30 AM
All I can think about is from Grand Theft Auto: Vice City - "In the future, there will be robots!" and "Man and Machine, living together, with bangs." (IIRC, of course)
Posted by: Naked Bunny with a Whip
|
May 22, 2009 11:31 AM
@dogmeatib #35:
Many people already suffer through this scenario. You've already sussed the solution: everybody's doin' it! ^_^
I'm not sure I understand what you're asking here. I do know that people already have to coexist with people they despise right now.
I don't see why. Look what I can do, compared to people even 50 years ago. I can access information from the other side of the planet in a matter of seconds. I can talk to people scattered around the globe simultaneously. I can watch live video from space. I can query thousands of sources of information at once, and I can get a report containing the results of hundreds of thousands of calculations in a few seconds. And I still complain about how slow things are.
This conflicts with the last question I quoted. If my thought processes are faster, the answers I get from other computations take subjectively longer to finish.
Anyway, I suspect that machine intelligences wouldn't be running millions or billions of times faster than organic ones, just because of the vast number of basic calculations that have to be done. It's not like adding two numbers together = a thought.
I don't see why it would. A minute would still seem like a minute to you. Unless you mean because you'd just stand there, twiddling your thumbs, waiting for everyone around you to keep up. That might drive someone nuts, but why would they do it? People already multitask like crazy.
Good questions, dogmeatib!
Posted by: GBM | May 22, 2009 11:35 AM
To be honest, the notion of achieving immortality through 'uploading' your consciousness has always struck me as kind of laughable. Even if one grants that the technology to model consciousness is possible, it seems to me that the act of uploading yourself will not grant you immortality, it seems like it would only create a clone of you that wasn't capable of dying.
The way I think of it is like this, say you've just used one of his brain uploaders (this one doesn't disintegrate your body) Do you think that you will be simultaneously conscious of both your computer-self and your biological self? It seems to me that given that your consciousness is a brain state, once you and the computer had disconnected there would be two persons, not one; moreover if the process of uploading yourself did involve being disintegrated, all you'd have done is killed yourself to build a person that had the same memories that you used to have.
Posted by: Muffin | May 22, 2009 11:37 AM
"A bit of a kook"? Come on. He may be intelligent (for all I know, anyway), but the guy's not "a bit of a kook", he's completely batshit insane.
Posted by: the rockridge brother | May 22, 2009 11:37 AM
"Every time I criticize these guys, I have to brace myself for another flood of hate mail."
This is great news - I for one find your "I get email" posts among my very favourites here.
Posted by: Knockgoats | May 22, 2009 12:12 PM
alexandre van der sande,
Let me admit the "Rapture of the Nerds" description isn't mine! Can't remember where I heard it though.
Posted by: zaardvark | May 22, 2009 12:14 PM
Just looking into Kurzweil's successful predictions; they're all short-term "no shit" predictions. Computers will store files embedded with images, animation, and sound. Ooooooooooh. Cell phones will get smaller, and become more popular. Ooooooooooooooh. Solid state drives will become more popular in some devices. OOooooooooooooooooooh!
His long-term predictions should not be lent any weight by his past "predictions". These are all things that industry people could have told you, without the Nostradamus schtick.
Posted by: Muffin | May 22, 2009 12:42 PM
"A bit of a kook"? Come on. He may be intelligent (for all I know, anyway), but the guy's not "a bit of a kook", he's completely batshit insane.
Posted by: lightning
|
May 22, 2009 12:56 PM
I'm glad I'm not the only one who thinks Kurzweil is the current equivalent of Madam Blavatsky -- nice sounding phrases so loaded with vague descriptions, undefined terms and unstated assumptions as to be meaningless.
Posted by: amphiox | May 22, 2009 12:56 PM
#9: Kasparov was beaten fair and square. He accepted those match conditions, having handily beaten the earlier version of Deep Blue one year earlier, under identical conditions (though in that match he lost one game and that was the first loss by a reigning world champ to a computer). He assumed the new version would play like the old version, prepared as such (having access to all the old version's games), and was surprised when the computer didn't play the way he expected. He came in overconfident and arrogant, and richly deserved to lose.
That said, Kasparov won the first game brilliantly, lost the second when the computer surprised him, then held the balance of play in the next three games but was unable to convert those small advantages to victories against the computer's precise endgame play, and then self-destructed in the last game, playing incredibly poorly and falling into an obvious and well known opening trap to boot. Any average chess master would have cleaned Kasparov's clock if he had played the same way against them as he did in that last game.
So Deep Blue certainly did not prove any convincing superiority of ability in winning that match. At best it demonstrated a rough parity with top level human play, and showed that it could beat the world champion when the human was having a bad day.
Today, though, the superiority of the best computer programs is obvious (though the magnitude of the difference still remains slight).
But the nature of chess, the kinds of mental abilities it requires, is in fact something computers do inherently well, while human brains do inherently poorly. Chess is hard for humans, which is why we admire the ability as much as we do, while it is easy for computers (relatively). So for a computer to outstrip us in an activity in which it is inherently well suited, and in which we are not, is not an impressive a feat as some might think.
Posted by: Demechrias | May 22, 2009 1:09 PM
Futurism, by nature, is woo-like: anyone trying to predict more than a few decades into the future cannot reliably use the scientific rigor we value so much. Thus, it tends to be more science fiction than science, a sort of exercise of the imagination. Thats not to say its useless, however, as any fan of science fiction can testify to. At least futurists will admit speculation and discuss the premises, unlike hard-core woo peddlers.
That said, I'm always somewhat critical of those who say what definitely will NOT be possible half a lifetime from now, those who see the obstacles of the present day as fundamentally insurmountable. With this attitude, how exactly does one envision five decades from now? Exactly the same, but our ipods are a bit smaller, our video games a bit crisper, our cancer treatments slightly more effective? To me, this way of thinking is just as bizarre as Kurzweil's, and far less imaginative.
Posted by: Brian X | May 22, 2009 1:09 PM
You know, I honestly can't take the Singularity ideas seriously at all. If such a thing was ever going to happen, one could easily argue that it already happened, sometime in the 19th century, either with the telegraph or the widespread adoption of railroads. Kurzweil sees a world like "The Matrix"; I see "My Life As A Teenage Robot", and given how little AI as a research field has panned out despite everything it's given us in pattern recognition and expert systems makes me think even XJ-9 is a little overly optimistic.
(Which is not to say research on AI has been worthless, but for what it's given us compared to its original aims, it seems to have more in common with alchemy than chemistry -- a source of ideas for other fields, but a dead end on its own.)
James Sweet #9:
Horseshit. Kasparov may be the greatest chess player in the world, but he's also not known to be entirely in touch with reality (among other things, he's a supporter of the Fomenko New Chronology), so his critical thinking skills are on record as being pretty horrible.
Personally, I believe that people accusing IBM of cheating are suffering from a very bad case of sour grapes poisoning. Chess is a brutally mathematical game -- ironically, for all the work that's been done in the AI community on computer chess it has very little to do with intelligence and much more with memory and search algorithms -- and it lacks the complexity that leaves, say, Go well out of the range of computability.
Posted by: Citizen of the Cosmos
|
May 22, 2009 1:11 PM
I doubt that there are very many people who actually don't believe progress in computer power or AI will continue. AI on our level seems entirely possible, if not inevitable.
Or will progress in these areas somehow just stop one day?
Posted by: BlackSun | May 22, 2009 1:15 PM
PZ, even you can be a douche. I agree with you 95% of the time. But I haven't seen one reasoned critique of Kurzweil. Just a bunch of sloppy ad hominems. He has a whole section in his book refuting the critiques of others, including some pretty big names like John Searle. I suppose you consider every word of that also to be more fluff and "junk science." Care to be a little more specific?
His predictions are a lot more detailed than what's been mocked here. And his ideas about the implications of double exponential growth seem to be unchallenged. Every day, he's further vindicated and his detractors look sillier. Still, futurism takes balls, especially near-futurism. So far, he's done brilliantly. How freaking easy for bystanders to leer and jeer and say "that was obvious." Well, obviously, it's 20-20 hindsight.
"He wants to become immortal." What an easy target you've picked. This is simple hate. Outside of religions, I've never seen such hatred. "How dare he try to become as GOD?!" Who can blame someone for trying? What, is he Icarus? Before man learned to fly, there were similar hoots of mockery. Kurzweil is the daVinci of today.
Calling something a "Kult" isn't an argument, merely a smear. Declaring that you expect to get hate mail doesn't mean you're right, just aware that other people passionately disagree with you. Have the courage to admit you don't know. He might be right about a few things, or mostly right or spectacularly wrong. Either way his work is too intriguing to ignore.
Posted by: Brian X | May 22, 2009 1:17 PM
Citizen of the Cosmos:
I think the current state of AI research is somewhat akin to someone trying to figure out the rules to Mornington Crescent. (If you understand why many people scoff at the term GNU/Linux, that's the problem in a nutshell -- there's a superstructure, but no core.) There simply isn't enough known about the nature of intelligence to do anything but nibble at the edges. As I said, that nibbling has been extremely productive, but the fundamental goal of AI simply doesn't appear to be reachable, at least not with what we know now.
Fifty or a hundred years from now, with a better knowledge of exactly how the brain *does* intelligence, who knows -- I could be proven wrong; it's not like there's a hard limit like the speed of light. But Kurzweil's Singularity assumes that AI research is unlimited, while practical experience shows that for the forseeable future, it's very limited indeed.
Posted by: Citizen of the Cosmos
|
May 22, 2009 1:29 PM
#75
There is progress being made in AI, and modelling neural networks, etc. We have not achieved human level intelligence, but there is absolutely nothing that would make such an achievement impossible. I don't expect it to happen within ten years, but I do expect it to happen well within 100 years. Our understanding of our own brains increase all the time, and therefore our chances of creating better models of neural networks. Computer power is increasing exponentially, which means one day soon, computers will surpass human brains in raw computing power.
Superhuman AI will be created one day, and assuming that computing power will continue to increase, and that a superhuman AI could improve on its own design... things might get really interesting. But there will be no more significant progress beyond where we are now, will there?
Posted by: JM Inc. | May 22, 2009 1:33 PM
Ugh. Every time i read about this guy now I want to say, "Hey, Ray, we all want to be cyborgs. We all want to have disease-proof, harm-proof, age-proof bodies and fault-proof brains. But that doesn't mean we all decide to join or build a technotopic robot cult in a certainly futile effort to get there."
Consensus science is a ruthless expunger of crap, and Kurzweil's crap, fortunately for us, comes nearly pre-expunged; self-marginalising, and intellectually bankrupt in the most obvious of ways. Unfortunately, on the other hand, that has never stopped people before.
Posted by: E.V. | May 22, 2009 1:33 PM
Right on cue, BlackSun seems to be a fanboi of Kurzweil. His quasi fourth sentence reads, "Just a bunch of sloppy ad hominems, " which he follows with... sloppy ad hominems. Irony - ain't it a bitch?
Posted by: Cameron | May 22, 2009 1:43 PM
From a complete layman's perspective here:
I read "The Singularity is Near" and I had one glaring issue with it. If we are able to upload a copy of our mind to a computer, or whatever they'll be called in the future, doesn't that mind immediately assume it's own "identity" by knowing it's a copy? It immediately would know it is different than the original mind and they would have different experiences, correct?
Posted by: Citizen of the Cosmos
|
May 22, 2009 1:45 PM
I was wrong. People really do believe there will be no significant progress in AI from now on. Fascinating.
Posted by: Cameron | May 22, 2009 1:47 PM
From a complete layman's perspective here:
I read "The Singularity is Near" and I had one glaring issue with it. If we are able to upload a copy of our mind to a computer, or whatever they'll be called in the future, doesn't that mind immediately assume it's own "identity" by knowing it's a copy? It immediately would know it is different than the original mind and they would have different experiences, correct?
Posted by: Naked Bunny with a Whip
|
May 22, 2009 1:57 PM
@Cameron #81: Yes, that's an issue for any nondestructive, discontinuous duplication process.
Posted by: Mushroom | May 22, 2009 1:58 PM
Kurzweil may seem kookish, but he backs up his ideas very well. It's hard to see how the future he envisages could fail to happen eventually, unless you believe conscious AI is impossible, or unless you think we'll destroy ourselves first.
dogmeatib:
No. Read this, from someone who thinks Kurzweil's predictions are mundane.
Posted by: BlackSun | May 22, 2009 2:09 PM
"fanboi?"
E.V.: argument, please?
As I said, he may be right, he may be partially right, or he may be wrong. Irony is a bitch, isn't it?
Posted by: BlackSun | May 22, 2009 2:13 PM
"fanboi?"
E.V.: argument, please?
As I said, Kurzweil may be right, he may be partially right, or he may be wrong. Apparently saying "not so fast" makes me a Kool-Aid drinker. So much for critical thinking, a casualty of yours and PZ's bum's rush to judgment. Get the pitchforks.
Irony IS a bitch, isn't it?
Posted by: Alan Kellogg | May 22, 2009 2:14 PM
#82
Now consider all the people who cannot, or will not, take into account the fact we are talking about copies. Not the original, but a copy. It is not you who would be uploaded to the computer, but another derived from you. And we're talking about a dynamic process, not a static object.
Now consider that the modern computer is more like a very sophisticated neuron instead of that blob of interconnected computing devices and complex organic interwebs we call a brain. Plus the fact that the basic hardware is fundamentally different and you can see a lot of difficulties with current ideas regarding AI.
BTW, the singularity started back around 10,000bc with the development of agriculture and the start of urban society. A Man created environment that had never existed before, placing evolutionary pressures on the species never possible before. Already we have made adaptations that have changed us, and we cannot see how we will be further changed in the generations to come.
Posted by: ihedenius | May 22, 2009 2:25 PM
After a surgery I noted how the nerves were cut off under the scar. It was creepy, might as well have run my finger over plastic. At another time taking a pain medication that seemed to shut all sensory nerves off, left me feeling like a huge cottonwad, I prefered the pain.
Point is, transfering to a computer mind, I think we will demand the full sensory input of a real body or we will, if not go insane, be very unhappy.
Also, what kind of a mind would it be without all the 'automatic' lower urges, fear, love, survival drive, would it be a mind at all (the limbic system ? not a neurolog). Maybe we need to understand ourselves more before we can hope to succede creating an AI.
Posted by: Brian X | May 22, 2009 2:28 PM
Citizen, #76:
Conscious AI is possible in the same sense that teleportation is possible -- there's no scientific barriers to it, but that doesn't mean it's physically possible based on what we know now. The simple fact is that you can make any assertions you like about what is and is not inevitable, but it's all contingent on having a reasonable mapping of what intelligence actually is. We do not have this yet. We're not even close -- psychology right now revolves primarily around using therapy to repair gaps in logical reasoning while using drugs to tweak neurotransmitters with the precision of sledgehammers. Models of intelligence and sentience are primitive, highly debatable (and debated), and to me it's the height of hubris to look at what AI has done and to say that it somehow shows that creating intelligence de novo is obviously possible. Compared to that, the old debate over top-down vs. bottom-up implementation seems more or less moot pending future advances.
As for raw processing power, IMHO it's a red herring. Processing power enables algorithms that had been impractical (I'm typing this on a laptop that needs a hardware MPEG-2 decoder to play DVDs, something nothing on the market now needs), but the algorithms have to exist first in order to make use of it. Again, I'm not saying strong AI is impossible, but you can't get there from here.
Posted by: Lepht | May 22, 2009 2:29 PM
oh, man, PZ, you have no idea how good it is to hear somebody criticise Kurzweil. i work with implantable technology - things that we actually do have today and can install, like RFID ampoules and neodymium magnetic sensors - and those dreamers are the bane of the field; they make all of us look like gibbering idiots with no research and no concrete technology, nothing but comic-book sci-fi fantasies. every time i suggest some tiny improvement to what we have, i get lambasted by Singularity followers wanting to be uploaded to Google or some such nonsense.
thanks for getting some criticism out there. the H+ community really does appreciate it.
Lepht
http://sapiensanonym.blogspot.com/
Posted by: Naked Bunny with a Whip
|
May 22, 2009 2:34 PM
Yep. Of course, it's possible the original is destroyed in the copying process, making the copy the only remaining copy. Alternatively, the "upload" process may be gradual, replacing cells in your brain with artificial processors over time. Then there was never an independently existing copy at all.
This all assumes, of course, that a human brain can be copied at all, and run on a different substrate. It may not be feasible. It may be that only a fundamentally different mental structure can be modeled effectively on artificial hardware.
It's fun to talk about this topic, but taking any strong positions with the available data seems pretty silly to me.
That's how I approach it. That's why I try to avoid slipping into dogmatism.
Posted by: Balk Onyx | May 22, 2009 2:42 PM
If this were the Matsumotoverse, this would end with a Japanese orphan with a poncho and sombrero exploding robot-Kurzweil's head with his trusty Cosmo Rifle, while a enigmatic Russian woman gazes on in silent approval.
(Alright, alright, I know it's much more likely that Kurzweil would get a robot body made entirely of Gouda cheese, and give a long impassioned and remorseful speech decrying the folly of his choice just before he dies heroically saving the boy's life on the Thousand Trillion Lactose Tolerant Hamster Planet. I just find it much more satisfying to imagine it the other way...is that so wrong?)
Posted by: bi -- IJI | May 22, 2009 2:46 PM
Shorter BlackSun:
They laughed at Galileo. (They also laughed at Bozo the clown, but let's ignore that.) Therefore, Kurzweil is the new Galileo.
This argument deserves a serious response.
-- bi
Posted by: The Very Reverend Battleaxe of Knowledge | May 22, 2009 3:21 PM
Black Sun:
John Searle is a bigger nut than Ray Kurzweil. If Searle didn't exist, Kurzweil would have found it necessary to invent him.Posted by: The Vicar | May 22, 2009 3:29 PM
@#48 (Jarrod):
The idea of the Singularity is a hypothetical point in the future where (in a nutshell) technology is so powerful that it erases fundamental rules with which we currently have to live. A few points:
- The hypothetical singularity is particularly characterized by artificial intelligences which can make themselves smarter, or at least develop other artificial intelligences which are smarter, leading to exponential growth (at least temporarily) in intelligence; I've seen a few people claim that this will be an outcome, not a cause, but it's usually given as a cause, and I think Kurzweil (who is one of the big Singularity writers) uses this as the specific way of identifying the Singularity, although frankly I find him so tedious I'm not even going to check Wikipedia to be sure.
- Due to the necessary deeper understanding of the nature of intelligence, it is theorized that if/when the singularity occurs, there will be effective immortality more or less limited only by the eventual heat death of the universe.
- Once the singularity has occurred, society will change so dramatically that existing history will be useless as a tool to predict future behavior, assuming that humanity isn't destroyed by robots or something as a result of it.
As you can see here, many people believe that some form of the Singularity is inevitable. Kurzweil believes it is not only inevitable, but coming soon enough that it will be within the current average person's lifetime.
For an interesting comic about the singularity, see the "Hob" storyline at Dresden Codak, which begins with the comic at http://dresdencodak.com/2007/02/08/pom/
Posted by: The Vicar | May 22, 2009 3:33 PM
@#48 (Jarrod):
The idea of the Singularity is a hypothetical point in the future where (in a nutshell) technology is so powerful that it erases fundamental rules with which we currently have to live. A few points:
- The hypothetical singularity is particularly characterized by artificial intelligences which can make themselves smarter, or at least develop other artificial intelligences which are smarter, leading to exponential growth (at least temporarily) in intelligence; I've seen a few people claim that this will be an outcome, not a cause, but it's usually given as a cause, and I think Kurzweil (who is one of the big Singularity writers) uses this as the specific way of identifying the Singularity, although frankly I find him so tedious I'm not even going to check Wikipedia to be sure.
- Due to the necessary deeper understanding of the nature of intelligence, it is theorized that if/when the singularity occurs, there will be effective immortality more or less limited only by the eventual heat death of the universe.
- Once the singularity has occurred, society will change so dramatically that existing history will be useless as a tool to predict future behavior, assuming that humanity isn't destroyed by robots or something as a result of it.
As you can see here, many people believe that some form of the Singularity is inevitable. Kurzweil believes it is not only inevitable, but coming soon enough that it will be within the current average person's lifetime.
For an interesting comic about the singularity, see the "Hob" storyline at Dresden Codak, which begins with the comic at http://dresdencodak.com/2007/02/08/pom/
Posted by: Michael Ralston | May 22, 2009 3:38 PM
The Vicar @94:
The problems with the form of Singularity you describe are thus:
It assumes AI is capable of indefinite exponential growth in intelligence. This seems unlikely to me - I anticipate an early semi-exponential phase, followed by an asymptotic approach towards a limit. Many Singularitan assumptions break if this is in fact what will happen.
I DO roughly agree with the immortality claim - brain uploads seem liable to happen eventually, and I find anyone who is concerned with the fact that an upload is just a "copy", but who isn't concerned about the fact that the vast majority of the atoms in their body will be replaced within a decade to be engaging in fuzzy thinking. And, of course, when minds are represented in digital data, backups will be eminently possible - data CAN be kept intact and in use indefinitely.
However, anyone trying to attach a date to it at this point in time is fundamentally making things up. We don't even know enough to know how much we need to learn.
The claim you make about societies so different that the past is no guide to the future ...
... well, we've been there. A lot. Hell, we're there now. Do you think anyone last century could have predicted some of the consequences of the internet?
How about the invention of the printing press? How about AGRICULTURE? This doesn't really say anything useful.
Posted by: The Vicar | May 22, 2009 3:46 PM
@#96: I wasn't defending the idea of the Singularity; I was responding to a post asking what the idea was in the first place. I am Singularity-agnostic; it's possible, I suppose, but whenever I try to imagine it actually happening, I remember how much hoopla was given to Vista. You just know that some big company which will be critical to brain transfers will turn out to have a big, gaping security hole that lets spammers transmit ads into your consciousness against your will, or use you as part of a botnet. In fact, if Microsoft is still around if/when the Singularity takes off, that will be the single largest obstacle to it actually occurring, forget about nuclear war.
Posted by: JD Cherry | May 22, 2009 3:50 PM
I actually authored the original criticism section on Ray Kurzweil's Wikipedia page, and it's good to see that it's growing. Ray Kurzweil and his followers drive me up the wall. They have absolutely no schooling in healthy scientific scpeticism. The word 'inevitable' is thrown around way too often. Kurzweil's wiki page is still far too long, (it lists his height and weight - what is he, a boxer?) but before I got there it was just hundreds and hundreds of lines of fawning praise, not unlike what you typically see in the media.
I think even the BBC reported once that "Machines will match man by 2029," in an article that repeated Kurzweil's claims verbatim with no analysis or contrary opinion. I was disappointed to see how many defenders Kurzweil had over at richarddawkins.net when that one was posted, by the way.
Funny how the many reporters will reverently repeat his claims about HAL-9000 arriving before many of us retire, but they tend to avoid his paranoid obsessive behaviour and his more outlandish claims. I haven't seen the "Jupiter-sized computers by 2100" or "Entire universe to be saturated with intelligent nanoparticles within several centuries" headlines, anyways. Hopefully this article (despite the negative connotations the author carries for some readers) will be a sign of a more critical attitude in the future.
Posted by: JD Cherry | May 22, 2009 4:03 PM
Oh, and furthermore when Kurzweil says immortality, he means, real actual eternal immortality. He believes that we will find a means to transcend the speed of light and send exponentially expanding nanoparticles to go out and disassemble the entire universe. All 100,000,000 galaxies. After the whole universe is "awake", it will decide to stop expanding and we'll all live indefinitely. The fact that this hasn't happened yet means that there are NO other intelligent, tool-using species in our great big beautiful universe, because if there were, the law of accelerating returns would have kicked in and we'd all have been swept up in the techno-rapture already. This is a very rigid system of beliefs.
In the short-term Kurzweil makes realistic predictions, but in the farther out you go, his predictions become implausible at a - dare I say? - exponential rate. As the article pointed out, many of his expectations have already failed spectacularly. He predicted by this year that the computer would cease to exist as a discrete object. Don't get me wrong - I'm no Luddite, and I love robots - but I really wish people didn't give Kurzweil so much press and credence.
Posted by: JD Cherry | May 22, 2009 4:08 PM
I forgot three zeroes on the galaxy number. My bad!
Posted by: uncle frogy
|
May 22, 2009 4:13 PM
everyone uses the term "AI" but I do not know what they mean by it. It sounds an undefined term at least as it is used in most popular uses. The Turing test is not referred directly. Intelligence is not defined either. Would we even recognize an AI if it was not a mirror of what we think our intelligence is?
We continue to be amazed at the intelligence of "less animals" on a regular basis seems like every time we look closely.
If you could "upload" your intelligence to some hardware does that not imply the you think it is a separate thing from the body not unlike a "soul"? Why would you think such a thing? As far a this singularity goes that is just a mental image from a human view point.
Inspirational thoughts and words and speculation are Ok if you don't take them so seriously. Everyone is right some of the time very few are right all the time.
Posted by: uncle frogy
|
May 22, 2009 4:16 PM
everyone uses the term "AI" but I do not know what they mean by it. It sounds an undefined term at least as it is used in most popular uses. The Turing test is not referred directly. Intelligence is not defined either. Would we even recognize an AI if it was not a mirror of what we think our intelligence is?
We continue to be amazed at the intelligence of "lower animals" on a regular basis seems like every time we look closely.
If you could "upload" your intelligence to some hardware does that not imply the you think it is a separate thing from the body not unlike a "soul"? Why would you think such a thing? As far a this singularity goes that is just a mental image from a human view point.
Inspirational thoughts and words and speculation are Ok if you don't take them so seriously. Everyone is right some of the time very few are right all the time.
Posted by: Steven Sullivan | May 22, 2009 4:20 PM
*sigh* someone feed this question into the singularity computer, please:
which of these internet species is more tiresome: Libertarians, Objectivists, or Transhumanists?
Posted by: BlackSun | May 22, 2009 4:49 PM
It's really time everyone calmed down. It seems the only position acceptable on this thread is that Kurzweil is a raving kook. If that weren't enough, let's throw in guilt by association: Libertarians and Objectivists. Horrors. Can the ___ party be far behind. I won't mention it because otherwise I'd be accused of violating Godwin's law.
It's my personal opinion that Kurzweil is like a modern DaVinci. I may be wrong about that. But before you go off half-cocked comparing him to Galileo or Bozo the clown in terms of being ridiculed, let me tell you what I meant.
DaVinci described the possibility of flight long before there was mechanical propulsion or the materials to build an airplane. But his drawings were sound and would have worked, if he had the necessary components.
Kurzweil is describing things that have not yet been discovered about the human brain. He's theorizing that all cognition is physical, and subject to reverse engineering. So obviously he's subject to ridicule? How many dualists are there on this site? If you're not a dualist, what's so strange about reverse engineering the brain to figure out how we think? And then augmenting our capabilities, upgrading our memory, installing web access, etc.? Is this really that outlandish? Can there not be some brain-machine interface short of full uploading? Come on, people...
I'm wondering if people would be more upset if he was right? Or if he was wrong? I've heard people debating Kurzweil becoming extremely distraught that all these people would be living longer lives, and "what would we do with them?" "How could we pay for them?" "There's too many people here already." "Death is what gives meaning to life, etc. etc." It's typical deathist rhetoric. They dig your grave for you and wonder why you're not happy to lie down in it.
I think most of Kurzweil's detractors are afraid he might be right and we might all live a lot longer. Not immortal, just maybe double or triple the current lifespan--in reasonable health. Who wouldn't take that deal? Mocking him won't change the path of discovery one bit. Nor will it prevent people from having their brains repaired or upgraded when the technology permits. You watch.
None of us knows the future, so why don't we just chill out and see what happens? Why the need to put constraints on human knowledge or progress, and accuse other scientists of kookery? I think it's an unscientific cop-out, a gross failure of imagination.
Posted by: Stu
|
May 22, 2009 4:54 PM
which of these internet species is more tiresome: Libertarians, Objectivists, or Transhumanists?
Oh, Libertarians, by far. At least the others are occasionally entertaining.
Posted by: The Very Reverend Battleaxe of Knowledge | May 22, 2009 5:12 PM
Perhaps Black Sun or some other upload enthusiast could tell us, after you've uploaded a complete analog of your brain, its current state, all of your memories, etc., into a computer so that there is a fully-functioning duplicate of you running on it...exactly What the Hell good does this do you, still stuck in a body that's going to die? So what if there's a copy of me running somewhere? I'm still stuck in here!
Posted by: BlackSun | May 22, 2009 5:23 PM
Reverend Battleaxe,
There's a late summer festival where they burn such 125-foot-tall straw men. You should go. I never said I was an "upload enthusiast" as you pejoratively inferred. I said "Can there not be some brain-machine interface short of full uploading?"
But in case you wondered, here's how it's supposed to work: You gradually replace parts of the brain with silicon or whatever other material can duplicate or assume neural activity. Pretty soon you have a complete functioning brain in silico that would still be perceptually you. Obviously this is theoretical, but that's how the theory goes. As you said, a mind-clone would clearly be a mind-clone, not you.
By the way, this has personal significance to me because my mom lost her mind ten years ago to Alzheimer's. If this had happened in, say, 2025, she could have gradually replaced parts of her disintegrating brain with prostheses.
It goes without saying that if you can help sick people this way, there's nothing to stop healthy people from getting the same implants. And the problem with that is what, exactly?
Posted by: JD Cherry | May 22, 2009 5:23 PM
Mr. Blacksun, I think that the problem most people have with Kurzweil is that extraordinary claims demand extraordinary evidence. He is completely certain of himself, and so many of his followers are so adamant that he is correct, and yet his evidence amounts to little more than a line on a 2-D graph. To echo Daniel C. Dennett, Kurzweil and his supporters massively underestimate the complexity of life. Our frustration is redoubled by all the positive attention he gets.
And if anyone is a "deathist", it's Kurzweil. I doubt any of us here are fundamentally opposed to the very idea of artificial intelligence, or even the possibility of eventual radical life extension, but Kurzweil's entire scheme is set up so that a rapture can sweep in and save him from death, right at the last minute. He's so obsessed with death that he's afraid to smile or speak beyond a monotone, in order to avoid "mechanical damage" to his body. His ideas have to be taken with a grain of salt, and with his disciples and too often the mainstream media that's not the case.
Posted by: Naked Bunny with a Whip
|
May 22, 2009 5:26 PM
No, of course not. It's an acknowledgment that there is nothing magical about meat.
Posted by: Stu
|
May 22, 2009 5:30 PM
It's really time everyone calmed down.
Fuck you. Who died and made you moderator?
It seems the only position acceptable on this thread is that Kurzweil is a raving kook.
Well, since he is a raving kook about many things, it is the only defensible one. But YMMV.
If that weren't enough, let's throw in guilt by association: Libertarians and Objectivists.
Noone did that. Your grasp of reality seems tenuous.
Can the ___ party be far behind. I won't mention it because otherwise I'd be accused of violating Godwin's law.
What the hell are you smoking?
Oh, and ___ does not exempt you from Godwinning the thread if it is obvious what you mean, moron.
It's my personal opinion that Kurzweil is like a modern DaVinci. I may be wrong about that.
You're entitled to your opinions, not to your own facts. He's been obvious with some predictions, and plain wrong in others. I'm sorry he's not infallible, but that's reality.
DaVinci described the possibility of flight long before there was mechanical propulsion or the materials to build an airplane.
Yes, because he looked up and saw birds, so he knew it was possible. I doubt anyone here will dispute that it will someday be possible to build something equivalent to the human brain. I firmly believe we will -- I also think that we won't be able to fully understand or predict how it will work.
But his drawings were sound and would have worked, if he had the necessary components.
Awesome. So does Kurzweil have blueprints for this AI? No?
Analogy fail.
Kurzweil is describing things that have not yet been discovered about the human brain. He's theorizing that all cognition is physical, and subject to reverse engineering.
Yes...
So obviously he's subject to ridicule?
Oh come on, take your strawman argument and shove it. You know that his predictions get bug-fuck crazy after that, and all you have left is arguing this. Knock it off. You're not fooling anybody.
what's so strange about reverse engineering the brain to figure out how we think?
Nothing, and not a soul here argued that (unless I missed something). You can put this one with the others.
I'm wondering if people would be more upset if he was right?
What the fuck? Are you now throwing the "you're just afraid" card?
I've heard people debating Kurzweil becoming extremely distraught that all these people would be living longer lives, and "what would we do with them?" "How could we pay for them?" "There's too many people here already."
What, those are not very, very valid questions?
"Death is what gives meaning to life, etc. etc." It's typical deathist rhetoric.
What the hell is a deathist? You just made that up, didn't you?
They dig your grave for you and wonder why you're not happy to lie down in it.
This is morbid, paranoid and reeks of a pathological persecution complex. Did you take your lithium this morning?
I think most of Kurzweil's detractors are afraid he might be right and we might all live a lot longer. Not immortal, just maybe double or triple the current lifespan--in reasonable health.
What makes you think that. What. The. Fuck. Makes. You. Think. That.
Who wouldn't take that deal?
I'm not sure I would.
Mocking him won't change the path of discovery one bit.
Who said so? Nobody, that's who. Strawman number eleventyzillion.
Nor will it prevent people from having their brains repaired or upgraded when the technology permits.
Who said so? Nobody, that's who. Strawman number... ah, whatever.
You watch.
This is really, really, important to you, isn't it? Do you often dream of that day, one noone will be surprised by, desperately running in circles for someone to yell "neener neener" at?
None of us knows the future
Numbnuts, you've just spent hundreds of words implying that Kurzweil does. Can you keep a coherent thought for more than two paragraphs?
so why don't we just chill out and see what happens?
That's the plan, yes. But we like to make fun of kooks along the way. Makes us live longer.
Why the need to put constraints on human knowledge or progress, and accuse other scientists of kookery?
Is anyone telling anyone not to research this stuff? How many more strawmen can you fit up your incoherent ass?
I think it's an unscientific cop-out, a gross failure of imagination.
I imagine your upload will fit on a thumbdrive. Does that count?
Posted by: Alan Kellogg | May 22, 2009 5:50 PM
Blacksun, daVinci knew what he was talking about.
Posted by: BlackSun | May 22, 2009 6:08 PM
Stu, so you're a deathist (no, I can't take credit for that neologism, unfortunately) AND you can't read English. Got it.
The hairless apes don't like it when you propose there could be an intelligence greater than their own. It's either Frankenstein or The Rapture. Couldn't just be better. I don't expect the next 40 years to be pretty as people deal with the arrival of life-extension, artificial personalities and cognitive enhancement. It's coming. The only questions left are when and to what degree? Kurzweil may or may not have made accurate predictions, or they may be off by some number of years.
As to the outcome, what's important to me is that he's got people thinking of the implications before the technology exists. In that sense, he doesn't have to be either a kook or a prophet. Just someone conducting thought experiments and asking questions about what it means to be human vs. machine. Whether or not he's having a mid-life crisis or is afraid of death is really irrelevant. Let's focus on what he's saying, not speculating on why he may or may not be saying it.
But over and thankfully out on this thread. Trying to have a rational conversation about Kurzweil here is like being an atheist on a fundie site. So mock away, kiddies.
Posted by: Brian X | May 22, 2009 6:17 PM
It's really time everyone calmed down. It seems the only position acceptable on this thread is that Kurzweil is a ... yada yada yada ... accused of violating Godwin's law.
So defensive. You're not familiar with SIWOTI syndrome, are you? In any case, Kurzweil's ideas have real-world effects -- people pinning their hopes on contingencies that can't be predicted. How do we know people's productive lives aren't being damaged by transhumanist messiah complexes?
It's my personal opinion that Kurzweil is like a modern DaVinci. I may be wrong about that. But before you go off half-cocked comparing him to Galileo or Bozo the clown in terms of being ridiculed, let me tell you what I meant.
DaVinci described the possibility of flight long before there was mechanical propulsion or the materials to build an airplane. But his drawings were sound and would have worked, if he had the necessary components.
And we could travel faster than light speed if negative energy existed. Like I said earlier about strong AI -- can't get there from here. In Da Vinci's time, the closest thing to the necessary motive power that existed were Chinese rockets (which he didn't have access to) and Heron's steam jet engine (which no one had ever figured how to do useful work with, and might have been forgotten in the 1300s). The term you're looking for is "left as an exercise", in its most twisted and sadistic sense.
Kurzweil is describing things that have not yet been discovered about the human brain.
How can he be describing them if they haven't been discovered? That's called speculation, not description.
He's theorizing that all cognition is physical, and subject to reverse engineering. [snip] Can there not be some brain-machine interface short of full uploading? Come on, people...
The idea is trivial. The implementation is incomprehensible with our current state of science. Like someone else said upthread, we don't even know what we don't know. Talking about anything that can't be directly extrapolated from what we can already do is meaningless until we can find out.
I'm wondering if people would be more upset if he was right? Or if he was wrong? I've heard people debating Kurzweil becoming extremely distraught that all these people would be living longer lives, and "what would we do with them?" "How could we pay for them?" "There's too many people here already." "Death is what gives meaning to life, etc. etc." It's typical deathist rhetoric. They dig your grave for you and wonder why you're not happy to lie down in it.
That's an awfully big straw man you're building there. You do realize the crows don't care so much after the first couple of days, right?
Immortality requires the repeal of entropy. Not only is that never going to happen, given the very nature of chaos the idea is inapplicable to our universe. Long virtual lives? Perhaps. But again, can't get there from here.
I think most of Kurzweil's detractors are afraid he might be right and we might all live a lot longer. Not immortal, just maybe double or triple the current lifespan--in reasonable health. Who wouldn't take that deal? Mocking him won't change the path of discovery one bit. Nor will it prevent people from having their brains repaired or upgraded when the technology permits. You watch.
Have you shut your critical thinking facilities off completely? That's the only possible explanation for the above paragraph, and it's no different from what religious fanatics say when they hang hell over our heads.
None of us knows the future, so why don't we just chill out and see what happens? Why the need to put constraints on human knowledge or progress, and accuse other scientists of kookery? I think it's an unscientific cop-out, a gross failure of imagination.
How positively postmodernist -- even New Age -- of you, giving a response barely distinguishable from garden variety Choprawoo. "Constraints on human knowledge" are based on observation of reality through a scientific process. None of Kurzweil's pet theories can be substantiated in any meaningful sense, and yet here you are defending them as if they're a fait accompli and he knows something the rest of the world doesn't.
Kurzweil reminds me at this point of Nikola Tesla or Edward Teller -- both brilliant idea men, but given more to promising people the impossible than to intellectual or scientific rigor. Kurzweil has done great work in electronic music and in handicap accessibility, but when he turned his thoughts to futurism, he hit a very hard, spiky wall known as the Peter Principle and hasn't looked back. I guess we're just lucky he isn't as dour and pessimistic as Bill Joy, who really does think we're headed to The Matrix, or worse.
Posted by: Stu
|
May 22, 2009 6:19 PM
Stu, so you're a deathist
Not at all, but thank you for playing.
Gratifying to see that after all your strawmanning and lying you've progressed to the final concern troll step: taking your ball and going home.
You will be sorely missed.
Posted by: FlyingSpaghettiTroll
|
May 22, 2009 6:34 PM
I'm not sure all what Kurzweil says, but I have little trouble seeing humanity transcend its self eventually. We've doubled our life span already. We can create artificial limbs that work off never impulses, our bones are already being replaced. The IBM blue brain project is trying to simulate the function of a human mind. They're making progress. We've got remote controlled rats, too. We've discovered an animal that recovers its base pairs.
Transhumanism as a whole isn't much of a question to me. If it comes from machines, so be it. Every few years your brain cells are slowly replaced, you aren't who you were, you're just a copy. I see little difference if some of it is replaced by identically functioning machines.
Trying to prepare for how we gradually will prolong our lives is stupid. We may just as well all die out. I leave near-immortality as a possibility worth considering, but one far off enough that I won't spend my life waiting for it.
-FST
Posted by: SocraticGadfly | May 22, 2009 7:07 PM
I like this comment by Davis:
I would spell "boring" as B-O-N-K-E-R-S.
As for his "cyborg" fantasies, OTOH, they really aren't that different than Dan Dennett's, at bottom line.
Of course, he's become more repetitive and regurgitative with each new book he writes.
Posted by: poke some slot | May 22, 2009 7:17 PM
As a data point, nutritional supplements are mostly neutral or harmful for you especially compared to a good diet without supplements, and Kurzweil takes tons of them. So he was wrong on that.
Posted by: poke some slot | May 22, 2009 7:50 PM
The problem I have personally with the Tech Rapture is the ambiguity. Where do you draw the line on immortal clutter of life? Do you make all the pets immortal too? Why not?
What about rapists and murderers? What about rapists and murderers who haven't been found until their brains get uploaded? Or are creatures of circumstances like the Somali pirates. What a tangle. I think the benevolent overlords will be like oh shit not this mess and let inefficient humans die off of old age, maybe with a stealthy nudge into sterility.
Posted by: Regalis | May 22, 2009 8:18 PM
I don't understand, is why people are so against the idea of extreme life extension. Kurzweil, and Aubrey de Grey, seem to have this reaction every time they are mentioned - either extreme antagonism or idolatry.
I'll ask the question here, since here hopefully I can get some discussion that doesn't revolve around some silly religiosity. What's wrong with trying for extreme life extension?
Posted by: Regalis | May 22, 2009 8:24 PM
I don't understand, is why people are so against the idea of extreme life extension. Kurzweil, and Aubrey de Grey, seem to have this reaction every time they are mentioned - either extreme antagonism or idolatry.
I'll ask the question here, since here hopefully I can get some discussion that doesn't revolve around some silly religiosity. What's wrong with trying for extreme life extension?
Posted by: uncle frogy
|
May 22, 2009 8:42 PM
naked bunny said No, of course not. It's an acknowledgment that there is nothing magical about meat.>>>
the mind does not exist without a fully functioning healthy brain meat. effect a part of it with chemistry, injury or disease and the mind changes . It seems to be the result of the interaction of all the "meat". it is the meat, the concrete physical reality, that is "magical" as you put it.
I thought that what was meant was uploading your mind to a machine and the mind would then be you forever in a machine. While you might be able to clone a mind unto a machine it would from that point be something else.
what do we mean by mind any way? Is it the human thought processes, the memories and learning, the personality? what about the emotional components which are tied rather strongly to the "meat" and memory to say nothing of the psychology of the individual. What would we upload to this computer.
fun speculation but like
the old saying many a slip between the lip and the cup.
Posted by: Michael Ralston | May 22, 2009 8:43 PM
Regalis: I haven't seen anyone indicate opposition to "extreme life extension". I've seen some people express skepticism about the possibility of such, and I've seen a lot of people declare that Kurzweil is a loon - but that's not the same thing, at all, as saying that what he wants would be bad.
It's just saying that the things he says don't connect up with reality.
Posted by: Gnimmel
|
May 22, 2009 8:52 PM
Douglas Hofstadter did an excellent talk on Kurzweil et al. in 2006 year ago:
Video:
http://www.singinst.org/media/tryingtomuserationally
Transcript:
http://www.acceleratingfuture.com/people-blog/2008/trying-to-muse-rationally-about-the-singularity-scenario/
He also mentioned the subject in an interview with American Scientist
Posted by: Otto
|
May 22, 2009 8:57 PM
I have known off Ray Kurzweil for a long time, so long that I quite forgot his original claim to fame. I was quite amused when I read about his ideas for eternal life, but I am convinced that his timetable is way off, by a factor of 5 to 10 I would guess.
Too late for me, sigh.
Will it be possible to transfer a "conciousness" into a "computer"?
Maybe.
Will it be worthwhile?
That depends on how this construct can interact with the world. It could be quite wonderful, it also could be a total nightmare.
I am happy to cheer Kurzweil on, he may be nuts, but it is a magnificent madness.
Futurists are optimists.
We had one in 1958 at the company where I worked and he made those wonderful predictions with absolute confidence, like
"By 1985 we will have fully functional automatic translation".
Hm.
Posted by: Gnimmel
|
May 22, 2009 9:12 PM
In general, this is not true.
Isn't this just the Ship of Theseus all over again? Suppose you collected the neurons as they were replaced and assembled them back into a functioning brain--now which one is the "real you"?
Posted by: andy | May 22, 2009 9:28 PM
I like the Iain M. Banks model - any sufficiently advanced A.I will want to keep us alive - for the lulz if nothing else...
Posted by: llewelly | May 22, 2009 10:08 PM
I don't understand why everyone at this site is so opposed to transpenism. Why don't you all admit you want 3 meter penes that can ejaculate 2 liters of semen?
Posted by: Otto
|
May 22, 2009 10:31 PM
llewelly,
I would like to live forever but not if I had to put up with a 3 meter penis.
Posted by: RamblinDude
|
May 22, 2009 10:47 PM
Well, it wouldn't be 3 meters all the time.
Posted by: Craig Holman | May 22, 2009 11:20 PM
I haven't read anything by Kurzweil in its entirety and I don't plan to. Life is too short to waste time on his kind of crap.
As far as I can tell, Kurzweil is interested in glorifying Kurzweil above all else. People with his kind of ego issues are very rarely worth paying attention to.
What I have read doesn't support the notion that he understands many things very deeply. He reminds me of an infomercial guy. Embarrassing.
I say this as someone who did his doctoral work in automated reasoning and who has given a lot of thought to issues related to AI. I think Kurzweil is clueless and just out for fame and money. People are foolish enough to give him both. What a con.
Posted by: WanderedIn | May 22, 2009 11:42 PM
This is just something I'd like to put out here; I don't expect to be noticed.
First off, I'm not a Singularitarian. Kurzweil seems a bit, um, well, kooky. Not inasmuch as he says that the Singularity could happen, as I feel that it's possible. Not inevitable, and definitely not soon, but possible. He's...well, I'd say really overoptimistic. But I'm kind of horrified by what forumgoers here did to Blacksun. As far as I could see, the critical responses to blacksun's posts were mostly just the equivalent of yelling, loudly, that KURZWEIL IS A KOOK, YOU ARE A TROLL, GO AWAY. After Blacksun left, Brian X gave a reasonably civil response, but Stu had already scared off the only real dissenting opinion. This is, like Blacksun said, alot like the fundie response to vocal atheists (and to be fair often the atheist response to vocal fundies on the atheists' forums, although comparing blacksun to a fundie would be a bit too cruel for my liking; Blacksun did, after all, put forward arguments, and responded at least coherently to questions)
But in any case, what we had here disappointed me; intelligent, non-fundie people, reacting about as spastically as Flavor Aid drinkers, because transhumies and such are designated internet punching bags.
I realize I am going to get called a troll, so I'd just like to say this: If you think the argument is irrational, take it apart. Don't just yell 'strawman'! or 'fundie!' or ask pointless rhetorical questions like "have you lost your mind?". All those do is let the opponent say, oh, these people don't actually have an argument, they just yell. As far as I can tell, you have arguments, pretty good ones, but don't use them, in favor of ridiculing the opponent. After all, ridicule is far easier, it gets a snort from allies, and it's quicker; it just doesn't convince anyone else.
Wandered In, futilely
Posted by: WanderedIn | May 22, 2009 11:46 PM
This is just something I'd like to put out here; I don't expect to be noticed.
First off, I'm not a Singularitarian. Kurzweil seems a bit, um, well, kooky. Not inasmuch as he says that the Singularity could happen, as I feel that it's possible. Not inevitable, and definitely not soon, but possible. He's...well, I'd say really overoptimistic. But I'm kind of horrified by what forumgoers here did to Blacksun. As far as I could see, the critical responses to blacksun's posts were mostly just the equivalent of yelling, loudly, that KURZWEIL IS A KOOK, YOU ARE A TROLL, GO AWAY. After Blacksun left, Brian X gave a reasonably civil response, but Stu had already scared off the only real dissenting opinion. This is, like Blacksun said, alot like the fundie response to vocal atheists (and to be fair often the atheist response to vocal fundies on the atheists' forums, although comparing blacksun to a fundie would be a bit too cruel for my liking; Blacksun did, after all, put forward arguments, and responded at least coherently to questions)
But in any case, what we had here disappointed me; intelligent, non-fundie people, reacting about as spastically as Flavor Aid drinkers, because transhumies and such are designated internet punching bags.
I realize I am going to get called a troll, so I'd just like to say this: If you think the argument is irrational, take it apart. Don't just yell 'strawman'! or 'fundie!' or ask pointless rhetorical questions like "have you lost your mind?". All those do is let the opponent say, oh, these people don't actually have an argument, they just yell. As far as I can tell, you have arguments, pretty good ones, but don't use them, in favor of ridiculing the opponent. After all, ridicule is far easier, it gets a snort from allies, and it's quicker; it just doesn't convince anyone else.
Wandered In, futilely
Posted by: WanderedIn | May 22, 2009 11:48 PM
Sorry about the double post...
My computer's a bit screwy sometimes.
Wandered In
Posted by: Ken Cope | May 23, 2009 12:12 AM
I don't think these ideas are terribly significant much earlier than The Firesign Theatre's I Think We're All Bozo's on This Bus, whose originality deserved precedence over the work of Minsky's friend Vernor Vinge, whose epic novella True Names is among the first fictional visualizations of "the Singularity," who gets far too little credit for the notion. Just about anybody on this blog would enjoy Vinge's recent Hugo winnner, Rainbows End. In personal conversation back in the 80's, Minsky scoffed at my enthusiasm for Wm. Gibson, along with other cyberpunks. Vinge, not Kurzweil, is the true owner of the concept and term "Singularity" which truly belongs to the storytellers, not the people who try and map reality onto constructs they've read about in not-all-that-popular science fiction. I've got to say that Hofstadter is the best referee and goalee these ideas have in academia these days. I'd have to say that as kooky as Kurzweil is, if it came to a choice between his view of reality and that of Penrose, I'd be stuck with Kurzweil. When it came down to it, when it was time for Tim Leary to choose between having his head removed and frozen and just going ahead with dying, just going ahead and dying didn't disturb him as much as the cryogenics folks did. My favorite author playing with these ideas is Rudy Rucker.
Posted by: Bryce | May 23, 2009 1:17 AM
I have some education in comp sci, has dabbled in AI a bit, and I've actually read The Singularity is Near. I'm satisfied that 90% of the Kurzweil-bashing I've read in the comments here is a mixture of misunderstanding and naked ad hominem.* The rest, I'll make a point of investigating further.
I think there are many potential singularity-like events, none of which are inevitable. Kurzweil's vision isn't as hard to achieve as many here argue. General AI research is in the doldrums right now, showing little sign of life. I was a bit underwhelmed by Wolfram Alpha. But if our tech-assisted meat brains can keep Moore's Law from going off the rails for a couple of decades, we'll have all the computational power we would need to simulate a brain directly. We'll also have powerful tools for dealing with the nano realm, and therefore should be able to collect as much information about the brain as we need to.
Now, if the bottom dropped out of Moore's Law right then, we'd be stuck with artificial intelligences about as smart as us, and who think about as quickly as we do. That's no help. But a few more doublings, and you have intelligences which can think on our problems much faster, though the quality of that thought may not be any better.
Set those minds to improving themselves, and soon you have better minds for dealing with the next set of problems.
I think Kurzweil is guilty of some overreach when he tries to extrapolate his logarithmic curve back before the invention of tools, and I see a lot of ways that the Singularity could fail to happen, or could wipe us out. But the predictions seem quite probable, given fairly mundane assumptions about the rate of improvement.**
Since some of you seem incapable of distinguishing between actual argument and simply spitting venom at the freaks, here's a list of assertions you could try to support, that if true would make things gravely difficult for the Singularity crowd:
* Moore's law will break down in the near future.
* Technological innovation will slow drastically, or even reverse, due to social disillusion, worldwide depression, or ecological catastrophe.
* Computer hardware cannot -- even in principle -- be used to simulate a human mind.
* The rate of technological progress is not exponentially increasing, as Kurzweil insists.
* Research into nanotechnology could be deemed dangerous, and a ban on it effectively enforced.
Actually, a few good points were made in regard to the second-to-last point.
Digression:
There is certainly room for doubt and skepticism. But not for the sort of blinding rage I see from some of the commenters. Even the resident blogger in chief isn't immune. Over here he claims to see zero difference between blind evolution, basic tool use, and a species that will soon master the ability to rewrite its own DNA. Why? Because a "futurist" claimed the opposite, and he really hates those guys.
It seems obvious to me that our species is unique, at least in the sense that it has more say over its form, function, and future than any species that has come before it. His response to that idea is dismissive:
That sounds less like a reflection on years of biological mastery, and more like unbridled nihilism. If everyone else is engaging in the backseat psychoanalysis of one Ray Kurzweil, I'll return the favor. The author has spent a lifetime watching mankind ravage the ecosystem, and Kurzweil pisses him off not just because optimism seems unwarranted, but because humanity isn't worthy of the predicted ascension.He wouldn't be entirely wrong.
* I mean, dear god, somebody compared him to Objectivists! That's just low.
** Okay, mundane given existing trends, which are themselves pretty outlandish and seemingly farfetched.
Posted by: Kagehi
|
May 23, 2009 2:03 AM
Frankly, the odds are far more likely we will see something like a Ghost in the Shell world, than a say Bubblegum Crisis. Yeah, I am an anime nerd. lol
The first premises development of more and more synthetic body systems, to replace damaged parts, to the point where the focus character (at least as represented in GIG3) ends up becoming one of the first "full body" transplants, with artificial systems. The "brains" are still, in most cases, organic, or semi-organic, in later models, relying on the "same" basic function as the current ones we have. The only big change that seems to have happened is the ability to pull such a brain case out of the machine, and put it in another, or, once the technology reached that point, for the consciousness to extend outside its normal bounds and do things it normally wouldn't, like copying itself into the, at that point, *far* more advanced computer networks. But, this might be (I am not so big a fan I know for sure), 100 years later, or more, from when the "initial" replacement was given to her. The premise in this one was a conflict, primarily, between the corporation's war development people, who liked to test the things in the city, where they would be confronted with less "lab condition" situations, and a group apposed to what the technology was being turned into. Sometimes, something from the war tech would slip into the consumer grade stuff, and all hell would break loose that way too (a bit like in one of the movies where people where "modding" their bots, and someone started slipping in "assassin" chips, or some such, into the third party markets for them, as a "test" of the technology). The result is a world with people that are from not, to partial, to complete body replacements, and even a few with artificial brains, but, the **AI** brains took decades, or longer, to get good enough to develop ghosts (real consciousness, with unpredictable capacities, instead of behaviors derived 'purely' from their programmed design specifications).
The second premises a scientific breakthrough. Mind, it has two separate time lines/technologies, depending on if its the original series, or the 2040 remake. The first version is traditional AI, in machine form, primitive, not too smart, simulation of organic systems, with "limited" learning. The smartest ones are almost human, and like the Terminator, have a living skin over them. The others are more mechanical, many not even made to look human. Both are sub-human intelligence, though the fully human looking ones specialize in "acting" human too, even though their learning capacity isn't too good, and they fail to interact properly in extreme cases. The second series made up a load of new "tech", that is totally improbable, including a liquid metal concept that later got "borrowed", or reinvented, in the second Terminator movie (hard to say who steals from who with these things, except when they make it obvious, like how Priss' band in the "original" series is called "The Replicants", as a nod to Bladerunner). In that second one, the AIs are all based on "simpler" forms of something that is like a processor core, with organic tissue around it, which was implanted in a real person, so it could "learn" the proper pathways to mimic a human mind. The result turned out to be so dangerous it was put in storage and frozen, then the company used "earlier", incomplete versions, to make their android servants. In this one, it was sort of the same, only it was a corporate cover up, by people that "also" owned and funded the local police force. The cover up: "Once these lesser semi-organics grew complex enough they would attempt to extend being programming constraints, in unpredicable ways, which 'usually' resulted in their liquid metal systems rearranging themselves to new configurations, developments of obsessions, and people dying as they attempted to continue to grow past those limits." Again, a vigilante group did what the police where, literally, not allowed or equipped to, and fought the things.
Yes, something is likely to happen, but I would say its closer to the GIS series than BGC, though, due to the unpredictability of such technologies, I wouldn't bet on there not being "some" stumbles into the fringes of the BGC style world too. The simplest reason being, its bound to be a lot easier to stuff a brain, and keep it alive, into a mechanical body, of some sort, than to recreate the brain itself, and there is at least "some" progress on keeping organs alive, without connection to a real body. And, more to the point, *any* technology that is likely to let you build a boomer (as the BGC androids are called), is going to let you replace bits of the human brain, to some extent, integrate things to it better, and eventually, *maybe* map the whole thing into it. Mind, its more likely to be a kind of black box affair. Something is "learning" to be you, without knowing it is, or you sensing it exists, then you die, and it keeps going. Any "integration" that may happen would extend the existing mind, make it reliant on the new configuration, and probably do "irreparable" harm to the mind in question, if "either" the organic brain, or the mechanical system, failed.
Posted by: Brian X | May 23, 2009 2:40 AM
* Moore's law will break down in the near future.
Unless quantum computing becomes practical, it will at some point. The limiting factors are a) how small can you etch pathways on chips (once we get to etching chips with far-UV, we'll know we're getting close), carrying capacity of ultrasmall circuit traces, and c) the effects of radionuclides in the substrate or etchants on signals.
* Technological innovation will slow drastically, or even reverse, due to social disillusion, worldwide depression, or ecological catastrophe.
Could happen, but that's at least as much a political problem as a technical one. Certainly the current problems with religious fanaticism are a major concern, as was the cold war 30 years ago.
* Computer hardware cannot -- even in principle -- be used to simulate a human mind.
As I said upthread, that's not a question we can answer right now. I would assume that based on current hardware, you'd need some kind of dynamically reprogrammable FPGA to simulate brain activity, but there's too much we don't know.
* The rate of technological progress is not exponentially increasing, as Kurzweil insists.
I don't think it's fair to place anything more than a very rough generalization on the mathematics of progress. Scientific advancement is a matter of contingencies, and not necessarily uniform. To take what I said about AI research upthread, AI has given computer science many useful items -- pattern recognition, neural networks, expert systems, etc. But it's never managed to make any real progress on actually capturing the nature of intelligence. I don't think a mathematical concept of progress even makes sense there.
* Research into nanotechnology could be deemed dangerous, and a ban on it effectively enforced.
Could happen. Probably won't.
Posted by: astrounit | May 23, 2009 3:42 AM
Ray Kurzweil once made a great digital keyboard.
Then he contracted singularitititis, and business flagged.
End of story.
Posted by: Bryce | May 23, 2009 4:16 AM
I think Kurzweil is predicting that the "lithography on a silicon wafer" technique will be done by 2019. He takes it as a given that we'll have a worthy successor to that approach by then, like quantum or optical computing. If people are talking about deploying these successors in 2015, the Singularity can proceed as scheduled. But if they're still only on the horizon by then, the future is less certain.
I bring up the possibility of an economic or environmental shock precisely because Kurzweil seemed too quick to dismiss it in his book. Some of his numerous charts tried to show that the curves were barely affected by major upheavals, wars, economic downturns, etc. That part made me somewhat suspicious, I'll admit.
Here I have to disagree. The question, in my estimation, is whether human consciousness is entirely a product of physical matter. If it is, then you can perform the processes of consciousness by simulating a human brain, at whatever level of detail is necessary for success.That level of detail may not be achievable, or we might not be able to perform the computations in a reasonable timeframe. But the only way to disprove it in principle is to invoke non-material phenomena, such as the soul.
I'm not sure what you're getting at with the FPGAs. Since any piece of hardware can be simulated on any other piece of hardware with sufficient memory to contain the simulation, there is no reason to invoke specific architectures until you're ready to worry about simulation speed.
Specialized hardware would run faster. But I don't think a fully reprogrammable FPGA would be ideal for the task. They have "neurons on a chip" that consist of, say, neuron-like circuits connected by synapse elements. You can lay out any pattern of neurons you like, to the limits of the hardware.
Now, neurons like those consist of a handful of transistors, and may not capture enough detail to describe conscious experience. But I think the principle is there.
So my answer to those who cite the failings of AI research is, we might not actually need to understand how it works. Just simulate this configuration of matter, and intelligence happens. Understanding the deep principles of consciousness would merely provide a shortcut, and reduce the amount of computation needed.
The other half of that equation is extracting the simulation data from a real brain. But that's not hard in principle. You just have to lose your ethical bearings long enough to perform a destructive scan with an electron microscope. It has already done on small amounts of neural tissue, according to Kurzweil.
Kurzweil's mathematics of technological progress aren't fully convincing to me. They tend to line up with my perceptions, but those perceptions have been influenced by various pundits saying that things are moving faster. I don't think progress is neat or easily measured, which dulls the force of Kurzweil's arguments.
But I still hope I get to be a robot.
Posted by: Stephanie W. | May 23, 2009 4:22 AM
andy @ #126-
I'm not familiar with Banks, but now I'm contemplating why we would be so funny to super advanced AI. I imagine that an easily-damaged, dependent naked ape who believes it is your god will present a pretty hilarious picture for Computo.
Essentially, our descendants will be the cats of the future. Look forward to the post-singularity human macros.
Posted by: Chris | May 23, 2009 5:04 AM
science fiction yes, religion no. Is dreaming a religion? is religion just some earthbound things made up by men? >Is the Bible science fiction?
Posted by: Chris | May 23, 2009 5:07 AM
science fiction yes, religion no. Is dreaming a religion? is religion just some earthbound things made up by men? >Is the Bible science fiction?
Posted by: Chris | May 23, 2009 5:14 AM
science fiction yes, religion no. Is dreaming a religion? is religion just some earthbound things made up by men? >Is the Bible science fiction?
Posted by: tochikoma | May 23, 2009 5:27 AM
The real tragedy is that if Kurzweil wrote actual science fiction we wouldn't be standing in this shitstorm trying to get our umbrellas open. Only in fiction can you throw out ideas as weird as some of Kurzweil's and make them palatable.
When you call yourself a futurist, suddenly you're seen as the apotheosis of douchebaggery. Conversely, people like Gibson, Sterling, Stephenson and others are accorded a bit more respect since many of the architects of our modern age read these authors and went out and built the things they read about.
Posted by: Kagehi
|
May 23, 2009 6:43 AM
The question, in my estimation, is whether human consciousness is entirely a product of physical matter.Invalid question. Modern neuroscience has been showing a strong inclination that "consciousness" as we understand the term, or more precisely, our awareness of our actions, and the justifications we apply to those actions, and post-hoc by products of a system that is **already** significantly slower than modern computer technology. In effect, we are able to perform faster than a computer in "most" circumstances only due to the fact that several layers of responses *begin* the process of performing some action, before we become "consciously" aware of having done so.
Think of it this way. Imagine that you have a "limited" internal network, and that network has 5,000 files on it. The files themselves are stored in such a manner that "parts" of the files are interconnected, where they share duplicate data, or where the data can be "interpreted" to produce near identical results, when processed as say, an MP3, instead of part of a picture, or part of a movie, or part of a text file. You make a request to get some file, the "server" opens up every single possible channel it can, and all the bandwidth possible, and delivers the first few bytes of the the 5,000 files, before you even "tell it" that you are looking for your vacation video. You pick "non-text", and 500 of those files drop out, while fragments of the remaining 4,500 get sent across. Each selection you make narrows the number of files you "need" to look at, increases the amount of each in memory, due to how they are stored, also increasing the amount of alternative data available, from which you can "reconstruct" a usable version of the final file, and decreases the amount you "need" to still retrieve. By the time you finally tell the OS in this fictitious network what file you "want", 80% of it is loaded, and 20% of it can be reconstructed, such that you can't tell sufficient difference as to care that its "not" the original file.
This of course does presume that the files themselves where "stored" in a way where the original is destroyed, and replaced with correlative associations, which allow them to be reconstructed, while using *less* actual data than the original (if it helps, imagine using a compression system that can cross link large file chunks in a way that lets it take 1k from file A, chunk 52, and simply tell file B to look at that "chunk", when it gets to its own chunk 234, which is identical, or near enough that the decompression system won't know the difference, or the playback won't show noticeable errors.
Mind, this example is a "simplification" of things. In reality, it would be more like 5,000 servers, each with 10,000 files, interconnected data, and the ability to "collectively" send the "final result" to your computer, in such a way that you are unaware of how it was reconstructed, where any of the parts where stored, or even if the original file a) exists in the system at all, or that b) what you are seeing "isn't" the original. And, as long as it was "close enough", likely no one would care "at all", until/unless it made a glaring mistake, like pulling the ending of Mall Cop up, in the middle of Watchmen, due to some miscue in what ever is used to reconstruct the data. And, this sort of miscue happens ***commonly*** in people, sometimes to an extreme, while at other times, just enough to leave you wondering what you original entered a room for, after being distracted by something "vaguely" related to the original intent.
Even from a purely practical standpoint, we can't even "recall" what we did, while doing it, with sufficient fidelity to claim that consciousness is "separate" from the hardware, so its really absurd to suggest it "exists" separate from it, even without all the evidence to suggest its not.
Posted by: Willem van Oranje | May 23, 2009 7:24 AM
"Some of Kurzweil's fellow futurists believe these superhuman computers will want nothing to do with us—that we will become either their pets or, worse yet, their food."
The Church is already fractioning.
Posted by: neurohacker | May 23, 2009 7:29 AM
how to Make Me immortal
1) Mind Uploading Home Page
http://www.ibiblio.org/jstrout/uploading/
see the Microtome Procedure
...(.~.).........
–oOO–(_)–OOo–
Bye for now and from the Past.
Have a nice day!
...neurohacker
nneurohackeratoms@gmail.com
Posted by: neurohacker | May 23, 2009 7:36 AM
how to Make Me immortal
1) Mind Uploading Home Page
http://www.ibiblio.org/jstrout/uploading/
see the Microtome Procedure
=======================================================================================================
2) Electron microscopy enters the picometer scale
http://en.wikipedia.org/wiki/Laser_microtome
========================================================================================================
3) Use a MedCosmCGHMaker
http://freshmeat.net/projects/medcosmcghmaker/
=========================================================================================================
to copy My BRAIN patterns into a HOLOGRAPHIC Base
=========================================================================================================
See
THE HOLOGRAPHIC BRAIN
http://homepages.ihug.co.nz/~sai/pribram.htm
Have a nice day!
Posted by: Knockgoats | May 23, 2009 11:09 AM
But if our tech-assisted meat brains can keep Moore's Law from going off the rails for a couple of decades, we'll have all the computational power we would need to simulate a brain directly. - Bryce
Bryce,
Like you I've read "The Singularity is Near, and I have a doctorate in AI (strictly "cognitive and computing sciences"), if we're comparing credentials.
As I've pointed out already, the computational power to simulate a brain would not imply the production of a superhuman AI able to redesign itself and hence increase its capabilities at high speed, as Kurzweil believes, even ignoring the fact that "Simulate a brain" is highly ambiguous: we don't know at what level the simulation would need to take place. Kurzweil, IIRC, assumes that simulation at the level of the synaptic network is sufficient, but we don't know that. If you needed to do it molecule by molecule, Moore's law would take a lot longer to reach the required level of raw computational power. However, let's assume he's right.
2) Simply "simulating a brain" wouldn't give you any useful capabilities. Whether a human brain does anything useful depends on it being connected to a body, and an external physical and social world, in the right ways, and educated for a couple of decades. Even then, the process can go drastically wrong.
2) Human intelligence is not limited by the capabilities of the individual human brain: pictures, texts, machines and interpersonal cooperation expand our problem-solving capabilities far beyond that. So a bigger or faster version of the brain, properly educated, would make a difference, but nothing like the difference Kurzweil naively assumes.
Posted by: albedo | May 23, 2009 3:31 PM
The only "futurologist" who was worth the name was Stanislav Lem. He hated the word very much (as it implies science where there is mostly fiction) and had a very critical look on his own "predictions", knowing the severe limitations of extrapolating the present into the future.
Posted by: The Vicar | May 23, 2009 6:43 PM
As far as Kurzweil's goal of conscious transfer/expansion/extension goes, this is not the real question. I have no doubt that consciousness is a byproduct of matter, but I don't believe that transfer/expansion/extension of consciousness is guaranteed.
The distinction is that consciousness is not the brain itself; consciousness is a byproduct of the brain, in the same way that "browsing the web" is not "a computer" but rather a thing which happens on computers. And if we're going to pursue that metaphor, then when you turn on the computer, you create a new instance of "browsing the web" -- to date, all such metaphorical computers operate in such a way that shutting them down destroys something unique, and there is no proof that this can be undone.
There are several difficulties:
- Transfer/expansion of consciousness, without merely making copies, more or less necessarily implies replacing parts of the brain one by one while it is still working. At either the functional level or the cellular level, the idea makes some pretty big assumptions about what is possible for the brain to survive.
If we're going to run with the "brain as computer" metaphor, you would need to replace every part of the computer one by one while it is running, without crashing it. While there are computers which are designed to handle this, there's no definite proof that the brain can do so. (Although the fact that neurons work in parallel is a point in favor of the possibility.)
- We don't really know where "consciousness" is in the brain, although we have some evidence about bits and pieces. It's possible that the whole brain takes part in some way which we have yet to trace. In that case, piecemeal replacement would necessarily kill consciousness, making the idea of consciousness transfer impossible.
- It is almost certainly possible to simulate a human brain all the way down to the molecular level at some speed. (Let's leave aside the questions of how much computing power would be required, and what fraction of the computing power available to humanity would be taken up.) In order for brain replacement to be feasible, one would have to find replacements which were more efficient than human biology. This is by no means guaranteed to be possible.
It may be the case that human consciousness requires actual neurons (or things which work the same way from all reasonable perspectives) to work. It may also be the case that the smallest thing which can behave like a neuron is actually a neuron. If those two assertions are true, then forget about replacing the brain cell by cell; it won't work without dropping the number of cells dramatically.
- The brain does not exist in a vacuum; it has inputs and is structured specifically to process data from those inputs. It is true that consciousness is adaptive to circumstances, but the brain itself is not. Brains are, in many respects, very fragile, and do not take easily to new input. (Yes, there is some evidence: there are people who are blind from birth, who can be "cured" by surgery. Having never developed the visual parts of their brains, they find it difficult -- and in some cases impossible -- to actually make sense of the input.)
While this is not an argument against consciousness transfer or extension -- you could always take a pair of cameras and map the signals into those of the optic nerves -- it does suggest that consciousness expansion will necessarily have some strict limits.
As for artificial intelligence, as opposed to tinkering with existing consciousness, it seems to be as far away as ever. So far, the Turing Test has only been passed under extremely specific and artificial circumstances. A real AI, of the type required for Singularity stuff, would have to pass what I think of as the Turing Test of Reality: if the AI were put in control of a remote-controlled robot body which was turned loose in society, could you distinguish between it and a human being with mechanical prostheses? So far, there exactly 0 AIs which even come close. In fact, the "expert systems" approach to AI -- where you don't care whether the result is conscious, but whether it is useful -- is producing most of the results of AI research, and the people trying to build consciousness are basically twiddling their thumbs and producing practically nothing.
Posted by: John Scanlon, FCD | May 24, 2009 12:26 AM
"...if the process of uploading yourself did involve being disintegrated, all you'd have done is killed yourself to build a person that had the same memories that you used to have."
Cuttlefish expressed the same thought: "...that thing, of course, would not be me."
And this matters because... you'd become separated from your immortal soul, I suppose. Yeah, that must be it.
And if it did not require disintegration to upload your memories, and you got to grow old and have long conversations, collaborations and whatever with your own cyborg copy and others, that would still be a bad thing because...?
The interesting questions of whether this sort of thing can happen in our lifetimes, the biology and physics and computer science and and economics questions, are being ignored by a lot of the commenters here*, in favour of these reflexively religious attitudes. This is supposed to be sceptical community, but also a thoughtful one.
*I'm only up to the first 50 or so comments, I'll catch up later.
Posted by: The Vicar | May 24, 2009 12:48 AM
@John Scanlon, #152:
The argument has gone over your head, here. The flaw is that the uploaded copy would be a copy, not the original.
Let's say that I start the process of uploading myself into the machine at t = 0, and the machine disintegrates me as it goes over a period of 10 seconds. As the machine works, I will experience whatever sensations accompany disintegration, but the uploaded copy will not. As I disintegrate, my consciousness -- which is a process, not an object -- will cease to exist, which is death. That there is a working backup copy somewhere else is irrelevant; it isn't really me.
In a way, it is the non-existence of the soul which is the objection, here. If there were a soul, then it is possible to imagine that the copy and the original share the same soul until the original is destroyed. But that does not happen; instead, the copy is a separate entity. When we talk about minds, rather than souls, disjoint-ness is a necessary objection.
Let's put it this way: suppose I invent a duplicator ray. Would you be willing to let me use the duplicator ray on you, and then vivisect you? Of course you wouldn't. The copy might behave just like you, but it is not you. It doesn't matter whether the copying process is displaced in time or not; the discontinuity makes you a separate entity.
Posted by: John Scanlon, FCD | May 24, 2009 1:54 AM
Yes, Your Reverence, a separate entity. However close a copy functionally, still not identical to the original in origin or potential. Of course that went over my head, because I'm stupid, right?
You and I and Ray Kurzweil, our bodies and brains are going to die (somewhat after our prime, assuming we're lucky), and because of the way life has evolved it's obviously technically problematic to upload the contents of our memories and personality to a more durable kind of hardware or software. But if it can be done once, there will be at least one immortal human. Whether other copies will be made from a lot more humans or only rubbed off from the first one - that's an interesting question. Whether vivisection as you describe would still be murder - that's not so difficult or controversial, or relevant at all.
Would anyone volunteer to be disassembled to create an immortal copy? I think we have the answer to that already. Not everyone would. But Ray would. I'd consider it, if I had evidence the technology worked. You have a problem with it? - well, it's unlikely to be compulsory.
Not that different from deciding whether to have children, really. Takes a bit of effort to bring them up, and it's not really immortality because even if they were clones, they would be different individuals from both parents. And we're going to die some day, whether we have progeny or not.
Also not that different from deciding whether to take a one-way trip to Mars, for example. There are parts of Earth that I wouldn't want to give up, but there are people who would volunteer in an instant. You going to tell them they're deluded? The US government may never have the balls to plan for one-way human space travel, but others will.
Posted by: Anton Mates | May 24, 2009 6:09 AM
Vicar,
Correct--assuming, of course, the machine's operation doesn't involve rendering you unconscious first, and thus unable to experience disintegration.
However, this doesn't demonstrate that the organic you is "you" and the uploaded you is not; a similar machine could do the same thing but disintegrate the uploaded copy after 10 seconds instead, and the organic you wouldn't experience that either. So the argument is symmetric as far as that goes.
By which standard, you die every night, and any time you get knocked out or put under general anesthesia. Surely, though, this is not how we normally define death; we think of that as a permanent, irrevocable loss of consciousness. If you lose consciousness--even for decades, as in a coma--and then wake up again, and your new consciousness bears sufficient resemblance to the old one, we do not consider you to have died at all.
So just have the machine knock you out first, copy and disintegrate you, and finally wake up your copy. The consciousness process has been restarted in a very similar state to where it was when it left off, just as if you went to sleep and woke up again. True, the process is now running on a different substrate, but that's true to some degree for anyone who loses and regains consciousness, given the intervening period of molecular turnover in the brain.
Certainly the copy is separate from the future you; the question is whether the copy has less claim to descend from the present you than the future "original" does.
Well, I wouldn't be willing to let you vivisect anybody, because that would be a nasty thing to do. That aside, I wouldn't be more or less troubled if you were planning to perform procedure X on the "original" me, vs. on my copy. (After the copying process I might care, but so might my copy!)
The "continuity=identity" principle fails in a quantum world, though. Copy or original, all your future selves are discontinuous from your present self; you cannot, even in principle, follow your component particles along continuous trajectories and determine which of your future selves they "really" end up in.
Put another way--a machine that "scans" you at point A and constructs a copy at point B is indistinguishable from a machine which teleports you from A to B, then leaves a copy in your original location. So how can one say that the person at point A definitely is or is not the original?
Posted by: The Vicar | May 24, 2009 11:48 AM
@Anton Mates:
Wow, you get a lot of stuff wrong in very quick succession.
Actually, this provides further evidence that the uploaded "me" is not really me. The fact that experiences are not shared between the two entities means they are not the same. And that means that as a means of obtaining immortality, such copies are useless.
Not really; when you are asleep, your consciousness does not go away; it merely enters a state in which is stops recording memories. See the book Counting Sheep: The Science of Sleep for a reasonably recent summary of the science behind what "unconsciousness" really means.
Well, then, let's take it to the next step: suppose I demonstrated that it is possible to make exact duplicates of things which once existed. Would you be willing to allow me to vivisect you if some third party you trusted promised to make a copy of you as you now exist at some point in the future? I'm guessing not. And even if you're weird enough to say yes, it doesn't mean you would become immortal, or even obtain any longevity.
Where on earth did you come up with that idea? Either you or one of your sources has been cherry-picking concepts from quantum theory. I really hate "I need an out, so quantum quantum quantum" because it leads to nonsense like this.
Except in rare circumstances which don't apply here, it is entirely possible to track nucleons -- otherwise chemistry would be impossible and our universe would be fundamentally different -- and it's nucleons which really provide identity.
(The key is mass: you cannot prove, even within reasonable doubt, that an electron you observe now is the same one as an electron at any point in the past because the mass of an electron is so small that Heisenberg's uncertainty principle -- which involves momentum, a quantity related to mass -- would allow its position to jump away while another election takes its place. Since any two fundamental particles with the same characteristics are identical, you would be unable to tell the difference. But the particles in the nucleus of an atom have masses roughly 2000 times that of an electron, and therefore experience much less uncertainty as to position. Therefore it is entirely possible to track them, at least to within a tolerance of certainty far beyond what would be necessary to validate claims of identity.)
There are two obvious flaws with this argument:
1) It has the unstated assumption that it is actually possible to teleport matter, as opposed to making a copy and deleting the original simultaneously. If we're going to deal with unproved hypotheticals like that, then we might as well go back to talking about souls.
2) The principles behind the machines would make it pretty clear which was the copy and which was the original; just because we don't know the information does not mean it does not exist. Or, to put it another way: suppose you have an antique vase. I take that vase into a locked room with no windows and make a copy which is identical as far as it is possible for you to tell. I then bring both of them back out. Just because you can't tell the difference between the two, it does not mean that one of them is not the original and one the copy.
Posted by: Anton Mates | May 24, 2009 5:35 PM
Vicar,
You're confusing two different identity issues. Of course the organic future Vicar and the uploaded future Vicar are not the same; they can have different thoughts and different experiences. There are most definitely two minds where there was one before.
But when you claim that “the uploaded 'me' is not really me,” you're claiming that the uploaded future Vicar is not the same being as the present you, and by implication that the organic future Vicar is the same being as the present you. Simply pointing to differences between your two future incarnations doesn't help that claim at all; what you need to do is show that one of them has a significant claim to continuity with your present mind that the other lacks.
It requires a weirdly expansive definition of consciousness to defend that through all states of sleep; but, as I said, you can also consider general anesthesia, or coma, or the few minutes after a sharp blow to the head. Or are you claiming that no living person ever experiences a temporary loss of consciousness?
Again, I don't like vivisection for anybody. But if you mean that you'd knock me out, “record” me, painlessly kill me, and then construct a future copy and wake it up—sure, that would be fine.
Not sure what you mean by this. If you mean that, barring medical advances, my future copy has about as much remaining lifespan to look forward to as my original version would have had, that's certainly true.
Unfortunately, quantum theory does appear to describe how the universe actually works. If you start making claims for identity which depend on identical systems being distinguishable...well, reality's against you.
I'm afraid not. It's possible to localize nucleons well--sometimes--but that's not at all the same as tracking individual ones.
And your “otherwise chemistry would be impossible” comment is pretty much exactly wrong. Thermodynamics depends on the fact that identical systems—whether electrons, nucleons, or entire molecules—are indistinguishable. If you wrongly treat them as distinguishable, your entropy calculations break because you're not counting your microstates correctly.
(Minor point: you really don't want to use an atomic nucleus as your example here. The nucleus is on the order of 10,000 times smaller than the atom, which more than makes up for nucleons' reduced position uncertainty. Individual nucleons occupy smeared-out orbitals within the nucleus, just as electrons do within the atom, and there's no way you could even begin to label a nucleon and track it over time. Your case would be better made with free nucleons in a vacuum, so let's go with that.)
The problem here is that you're starting from the wave function of a single particle, and saying that if the particle is massive, the measured position associated with that wave function tends to have a low uncertainty—which is, in general, correct. But we're not talking about a a single particle (or system), we're talking about two identical ones. And you can't just stack the wave functions of the individual particles to get the wave function of the pair. Once you do construct the correct combined wave function, there's no way to disentangle it into components definitely associated with each particle; hence, they're indistinguishable.
Beyond that, it's not possible to track them to arbitrary precision without destroying the behavior of the system, any more than you could do so for a set of electrons. To say you can track them “within a tolerance of certainty” is simply to say, “And after I've marked enough points along their trajectory, I have to give up and can't worry about whether teleportation occurred in this particular brief span of time.” As far as establishing continuity goes, that's worthless.
If we're not going to deal with unproved hypotheticals, why on earth did you invoke a “duplicator ray” in the first place?
Furthermore, you're missing the point of indistinguishability. If we can make a machine which takes an object here and constructs an exact copy there--and by your own assumption, for this discussion, we can--then the very same machine can be viewed as teleporting the object from here to there and leaving a copy in its place. There is no way, empirically, to distinguish the two cases. You can say "Well, I don't believe in teleportation, so this machine must actually be constructing a copy at a distance," but you can also say, "Well, I don't believe in unguided natural law, so magical elves must be guiding every particle to its destination." You can say just about anything.
Begging the question. The problem here is that the same principles are behind both machines. They're indistinguishable, remember?
And just because you can't detect any effects of the existence of the soul, doesn't mean it's not there—but why believe in it at that point?
If no conceivable test can distinguish between the “original” vase and the “copy,” then there's no reason to assign those labels to either.
Posted by: Anton Mates | May 24, 2009 11:35 PM
I don't think I said this as clearly as I could have, so just to go into more detail:
Suppose you have a massive particle 1 at point A, and an identical massive particle 2 at point B, and the points are far apart. By itself, particle 1 would have a wave function P_1(x), whose magnitude peaks sharply for x near point A and is almost zero elsewhere. Particle 2 would have a wave function P_2(x), whose magnitude peaks sharply for x near point B and is almost zero elsewhere. So, just as you say, you can be sure that Particle 1 would be at (or very close to) point A in the first case, and that Particle 2 would be at (or very close to) point B in the second case.
So far, so good. Now what happens to the combined wave function when you have both of these particles? Up to a normalizing factor, it becomes [ P_1(x_1) * P_2(x_2) ] +/- [ P_1(x_2) * P_2(x_1) ], where x_1 and x_2 are the positions of particles 1 and 2, respectively. (The plus or minus in the middle depends on whether the particles are bosons or fermions.)
The magnitude of this combined wave function peaks sharply for x_1 near A and x_2 near B, but also for x_1 near B and x_2 near A. Thus, even though you can be quite certain that each of the two points is occupied by one particle, it's impossible to say with any certainty which particle is at which point. Each particle is equally likely to turn up at either point. Hence, their individual identities have no physical meaning.
Posted by: melior | May 28, 2009 10:00 PM
Without taking anything away from the genius of Turing, the much ballyhooed Turing Test is an abominably poor substitute for a well-defined, rigorous methodology for distinguishing "artificial intelligence" from near misses. In fact, the only things it has going for it are that previous attempts were even worse, and no one has come up with a better one... yet.
I would hazard that by the time our understanding and technology advance enough to produce an actual "artificial intelligence" we will almost certainly have a more well-founded definition, along with a method of determining that the definition has been successfully met that involves far less hand-waving. In fact, it's hard to see how we will get there until we do.
Posted by: Alex | May 28, 2009 10:12 PM
"You know what I think of him, and the reporter interviewed me for a short comment, too."
He interviewed you? It seems more like he quoted something you wrote back in February from your blog. That's hardly interviewing you, but whatever.
Posted by: Elwood Herring | May 29, 2009 7:20 AM
Well, all I can say to all this is: "There's only one thing worse than being talked about, and that's NOT being talked about".
I bet Kurzweil is actually loving all the attention he's getting right now.
Posted by: John Morales | May 29, 2009 7:38 AM
Elwood H, indeed! :)
As to the topic, I give RK high marks* for his aspirations, but far lower ones for his expectation of their achievability (in the short to medium term, anyway, and certainly in their ultimate instantiation).
Purely subjectively, I consider true AI today is akin to where heavier than air flight was in Da Vinci's day, in terms of required accomplishment - it's plausible, but how it might be achieved I think is still speculatory and will require significant progress.
I personally will be impressed if I live to see it (DOB 1960).
--
* I note that he's more knowledgeable and smarter than I, in my estimation, to put my opinion in perspective.
Posted by: James Redford | May 29, 2009 8:43 AM
For someone who prides himself on his alleged rationality, you sure do make a habit of engaging in irrationality, Prof. P. Z. Myers. Just in this post by you, you have in a number of instances engaged in the logical fallacy of ad hominem attack. For example, your invectives of "kook," "kult," "technobabble," and so forth. Such logically fallacious behavior by you is particularly odd if you were in the right, as this is quintessential behavior of those who remonstrate against a particular position but who cannot marshal a reasoned argument against it, and hence who are left with logically fallacious handwaving exercises in an attempt to disparage the position they're against.
If Ray Kurzweil's position rests on something which violates then known laws of physics, then a legitimate refutation would be to show where his position violates the know laws of physics. And if a violation of the known laws of physics cannot be demonstrated, then there exists no legitimate scientific grounds in which to regard Kurzweil as a "kook" or engaging in "technobabble."
I haven't studied Kurzweil's books, but from what I have seen of his position, nothing about it that I've come across violates physical law.
Physically speaking, Kurzweil's main point that humans will be able to make themselves immortal is perfectly achievable. The idea that there is something special about human-level consciousness such that it cannot be created artificially unavoidably rests on the mystical position of vitalism: in effect that human-level sapience amounts to some type of soul that is beyond the study and description of science.
Moreover, the known laws of physics apodictically require mankind to take over all matter in the the universe, increasing their intelligence without limit, becoming omniscient, omnipotent and omnipresent at the end of time: as God has been proven to exist based upon the most reserved view of the known laws of physics. For much more on that, see Prof. Frank J. Tipler's below paper, which among other things demonstrates that the known laws of physics (i.e., the Second Law of Thermodynamics, general relativity, quantum mechanics, and the Standard Model of particle physics) require that the universe end in the Omega Point (the final cosmological singularity and state of infinite informational capacity identified as being God):
F. J. Tipler, "The structure of the world from pure numbers," Reports on Progress in Physics, Vol. 68, No. 4 (April 2005), pp. 897-964. http://math.tulane.edu/~tipler/theoryofeverything.pdf Also released as "Feynman-Weinberg Quantum Gravity and the Extended Standard Model as a Theory of Everything," arXiv:0704.3276, April 24, 2007.
Out of 50 articles, Prof. Tipler's above paper was selected as one of 12 for the "Highlights of 2005" accolade as "the very best articles published in Reports on Progress in Physics in 2005 [Vol. 68]. Articles were selected by the Editorial Board for their outstanding reviews of the field. They all received the highest praise from our international referees and a high number of downloads from the journal Website." (See Richard Palmer, Publisher, "Highlights of 2005," Reports on Progress in Physics website.)
Reports on Progress in Physics is the leading journal of the Institute of Physics, Britain's main professional body for physicists. Further, Reports on Progress in Physics has a higher impact factor (according to Journal Citation Reports) than Physical Review Letters, which is the most prestigious American physics journal (one, incidently, which Prof. Tipler has been published in more than once). A journal's impact factor reflects the importance the science community places in that journal in the sense of actually citing its papers in their own papers. (And just to point out, Tipler's 2005 Reports on Progress in Physics paper could not have been published in Physical Review Letters since said paper is nearly book-length, and hence not a "letter" as defined by the latter journal.)
See also the below resource for further information on the Omega Point Theory:
Theophysics: God Is the Ultimate Physicist http://geocities.com/theophysics/
Tipler is Professor of Mathematics and Physics (joint appointment) at Tulane University. His Ph.D. is in the field of global general relativity (the same rarefied field that Profs. Roger Penrose and Stephen Hawking developed), and he is also an expert in particle physics and computer science. His Omega Point Theory has been published in a number of prestigious peer-reviewed physics and science journals in addition to Reports on Progress in Physics, such as Monthly Notices of the Royal Astronomical Society (one of the world's leading astrophysics journals), Physics Letters B, the International Journal of Theoretical Physics, etc.
Prof. John A. Wheeler (the father of most relativity research in the U.S.) wrote that "Frank Tipler is widely known for important concepts and theorems in general relativity and gravitation physics" on pg. viii in the "Foreword" to The Anthropic Cosmological Principle (Oxford: Clarendon Press, 1986) by cosmologist Prof. John D. Barrow and Tipler, which was the first book wherein Tipler's Omega Point Theory was described. On pg. ix of said book, Prof. Wheeler wrote that Chapter 10 of the book, which concerns the Omega Point Theory, "rivals in thought-provoking power any of the [other chapters]."
The leading quantum physicist in the world, Prof. David Deutsch (inventor of the quantum computer, being the first person to mathematically describe the workings of such a device, and winner of the Institute of Physics' 1998 Paul Dirac Medal and Prize for his work), endorses the physics of the Omega Point Theory in his book The Fabric of Reality (1997). For that, see:
David Deutsch, extracts from Chapter 14: "The Ends of the Universe" of The Fabric of Reality: The Science of Parallel Universes--and Its Implications (London: Allen Lane The Penguin Press, 1997); with additional comments by Frank J. Tipler. Available on the Theophysics website.
The only way to avoid the Omega Point cosmology is to resort to physical theories which have no experimental support and which violate the known laws of physics, such as with Prof. Stephen Hawking's paper on the black hole information issue which is dependent on the conjectured string theory-based anti-de Sitter space/conformal field theory correspondence (AdS/CFT correspondence). See S. W. Hawking, "Information loss in black holes," Physical Review D, Vol. 72, No. 8, 084013 (October 2005); also at arXiv:hep-th/0507171, July 18, 2005.
That is, Prof. Hawking's paper is based upon empirically unconfirmed physics which violate the known laws of physics. It's an impressive testament to the Omega Point Theory's correctness, as Hawking implicitly confirms that the known laws of physics require the universe to collapse in finite time. Hawking realizes that the black hole information issue must be resolved without violating unitarity, yet he's forced to abandon the known laws of physics in order to avoid unitarity violation without the universe collapsing.
Some have suggested that the universe's current acceleration of its expansion obviates the universe collapsing (and therefore obviates the Omega Point). But as Profs. Lawrence M. Krauss and Michael S. Turner point out in "Geometry and Destiny" (General Relativity and Gravitation, Vol. 31, No. 10 [October 1999], pp. 1453-1459; also at arXiv:astro-ph/9904020, April 1, 1999), there is no set of cosmological observations which can tell us whether the universe will expand forever or eventually collapse.
There's a very good reason for that, because that is dependant on the actions of intelligent life. The known laws of physics provide the mechanism for the universe's collapse. As required by the Standard Model, the net baryon number was created in the early universe by baryogenesis via electroweak quantum tunneling. This necessarily forces the Higgs field to be in a vacuum state that is not its absolute vacuum, which is the cause of the positive cosmological constant. But if the baryons in the universe were to be annihilated by the inverse of baryogenesis, again via electroweak quantum tunneling (which is allowed in the Standard Model, as baryon number minus lepton number [B - L] is conserved), then this would force the Higgs field toward its absolute vacuum, cancelling the positive cosmological constant and thereby forcing the universe to collapse. Moreover, this process would provide the ideal form of energy resource and rocket propulsion during the colonization phase of the universe.
Prof. Tipler's above 2005 Reports on Progress in Physics paper also demonstrates that the correct quantum gravity theory has existed since 1962, first discovered by Richard Feynman in that year, and independently discovered by Steven Weinberg and Bryce DeWitt, among others. But because these physicists were looking for equations with a finite number of terms (i.e., derivatives no higher than second order), they abandoned this qualitatively unique quantum gravity theory since in order for it to be consistent it requires an arbitrarily higher number of terms. Further, they didn't realize that this proper theory of quantum gravity is consistent only with a certain set of boundary conditions imposed (which includes the initial Big Bang, and the final Omega Point, cosmological singularities). The equations for this theory of quantum gravity are term-by-term finite, but the same mechanism that forces each term in the series to be finite also forces the entire series to be infinite (i.e., infinities that would otherwise occur in spacetime, consequently destabilizing it, are transferred to the cosmological singularities, thereby preventing the universe from immediately collapsing into nonexistence). As Tipler notes in his book The Physics of Christianity (New York: Doubleday, 2007), pp. 49 and 279, "It is a fundamental mathematical fact that this [infinite series] is the best that we can do. ... This is somewhat analogous to Liouville's theorem in complex analysis, which says that all analytic functions other than constants have singularities either a finite distance from the origin of coordinates or at infinity."
When combined with the Standard Model, the result is the Theory of Everything (TOE) correctly describing and unifying all the forces in physics.
Posted by: John Morales | May 29, 2009 8:51 AM
James Redford, written like a true fanboi.
An exemplar of scientism, is what you are.
Posted by: /:set\AI | May 29, 2009 12:50 PM
Ray made Newsweek look like fools- especially about wearable computers/ no keyboards- my god this Newsweek idiot is entirely ignorant of iphone/ipod app explosion that has been happening for over a year! he obviously hides in his office and never looks at the people outside- all using ipod touches and iphones- networked computers in their pockets without keyboards
it was like if someone wrote an article today wondering why haven't yet had a black president (*___*)
come on! are all criticisms of Singularitarianism/ Transhumanism this foolish and oblivious to reality? one has to wonder if such articles are actually written by transhumanists to make the opposition look retarded- if so it works brilliantly
Posted by: Hank Roberts
|
May 29, 2009 1:09 PM
Kurzweil has it backwards.
In Singularity, computer uploads into you.
Please to lean closer to screen and hold very still.
This won't hurt a bit.
Posted by: James Redford | May 29, 2009 2:08 PM
Hi, John Morales. You wrote:
""
James Redford, written like a true fanboi.
An exemplar of scientism, is what you are.
""
That's two instances of the logical fallacy of ad hominem attack by you. Such logically fallacious behavior by you is particularly odd if you were in the right, as that is quintessential behavior of those who remonstrate against a particular position but who cannot marshal a reasoned argument against it, and hence who are left with logically fallacious handwaving exercises in an attempt to disparage the position they're against.
The only way to avoid the conclusion that the Omega Point exists is to reject the known laws of physics (i.e., the Second Law of Thermodynamics, general relativity, quantum mechanics, and the Standard Model of particle physics), and hence to reject empirical science: as these physical laws have been confirmed by every experiment to date. That is, there exists no rational reason for thinking that the Omega Point Theory is incorrect, and indeed, one must engage in extreme irrationality in order to argue against the Omega Point cosmology (as you amply demonstrate).
Additionally, we now have the quantum gravity Theory of Everything (TOE) correctly describing and unifying all the forces in physics: of which inherently produces the Omega Point cosmology. So here we have an additional high degree of assurance that the Omega Point cosmology is correct.
Bear in mind that Prof. Frank J. Tipler's Omega Point Theory has been published in a number of the world's leading peer-reviewed physics journals.[1]
Out of 50 articles, Prof. Tipler's 2005 Reports in Progress in Physics paper--which presents the Omega Point quantum gravity Theory of Everything--was selected as one of 12 for the "Highlights of 2005" accolade as "the very best articles published in Reports on Progress in Physics in 2005 [Vol. 68]. Articles were selected by the Editorial Board for their outstanding reviews of the field. They all received the highest praise from our international referees and a high number of downloads from the journal Website." (See Richard Palmer, Publisher, "Highlights of 2005," Reports on Progress in Physics. http://www.iop.org/EJ/journal/-page=extra.highlights/0034-4885 )
Reports on Progress in Physics is the leading journal of the Institute of Physics, Britain's main professional body for physicists. Further, Reports on Progress in Physics has a higher impact factor (according to Journal Citation Reports) than Physical Review Letters, which is the most prestigious American physics journal (one, incidently, which Prof. Tipler has been published in more than once). A journal's impact factor reflects the importance the science community places in that journal in the sense of actually citing its papers in their own papers. (And just to point out, Tipler's 2005 Reports on Progress in Physics paper could not have been published in Physical Review Letters since said paper is nearly book-length, and hence not a "letter" as defined by the latter journal.)
For the fuller details on these matters, see my previous post above.
-----
Note:
1. While there is a lot that gets published in physics journals that is anti-reality and non-physical (such as string theory, which violates the known laws of physics and has no experimental support whatsoever), the reason such things are allowed to pass the peer-review process is because the paradigm of assumptions which such papers are speaking to has been declared, and within their declared paradigm none of the referees could find anything wrong with said papers. That is, the paradigm itself may have nothing to do with reality, but the peer-reviewers could find nothing wrong with such papers within the operating assumptions of that paradigm. Whereas, e.g., the operating paradigm of Prof. Tipler's 2005 Reports on Progress in Physics paper is the known laws of physics, i.e., our actual physical reality which has been repeatedly confirmed by every experiment conducted to date. So the professional physicists charged with refereeing this paper could find nothing wrong with it within its operating paradigm, i.e., the known laws of physics.
Posted by: John Morales | May 29, 2009 8:48 PM
James @167:
Actually, I made no argument, merely expressed opinion. There is no logical fallacy where there is no argument.Not only does your reiteration of your original post strengthen said opinion, but I hereby append a further opinion: You either misunderstand what an informal logical fallacy is, or you're being disingenuous.
Spoing!PS I refer you to my #162, where I express my own opinion on the issue.
Posted by: Rorschach
|
May 29, 2009 9:24 PM
fanboi @ 167,
So I ask you then,if Tipler's numbers voodoo somehow "proves" god,what exactly does that change on Earth? Is he going to answer prayers for once now that we've found him out? Going to work some miracles is he? Cure the sick?Give creationists their brains back?
Because if not,then this whole mumbojumbo is utterly irrelevant.
Posted by: James Redford | May 30, 2009 12:57 PM
John Morales, your statement in post #168 is factually incorrect, as with your previous logically fallacious ad hominem attacks upon me. You have displayed quite a yen for being irrational.
Merriam-Webster's Collegiate Dictionary, Eleventh Edition defines ad hominem as "(1) appealing to feelings or prejudices rather than intellect; (2) marked by or being an attack on an opponent's character rather than by an answer to the contentions made."
Ad hominem attacks are in the category of argumentation, since even when the full syllogism isn't given, it is implied. Your point in personally attacking me was to imply that my position is incorrect.
Posted by: John Morales | May 31, 2009 4:26 AM
You're really dense, aren't ya? :)I called you a fanboi, and an exemplar of scientism.
That's describing my impression of you based on your comment, but it's not attacking your position in any way.
NB: ad hominem is merely Latin for "to the person" (which my comment was indeed); however, you wrote about "the logical fallacy of ad hominem", which would be denoted as argumentum ad hominem. Clearly, you can't distinguish between a fallacious argument and a statement of opinion.
Really. Since you claim such exists, what, pray tell, is my implied syllogism?Thick as a brick, is what you are, and with delusions of competence to boot, as evidenced by your misapprehension.
Posted by: Isam | June 10, 2009 5:58 PM
I have read all of Kurzweil books and honestly his predictions
all have come true but only ten years later than he predicted.
As far as the emergence of an AGI is concerned, I don't believe we will get there in the foreseeable future, but we will get close to it within about twenty years to 40 years.
With regard to the mind uploading of a person, this could also be possible, but we never get all of the person's subtle specifics and that uploaded software is similar to the person but definitely not the person.
Posted by: Singularity Hub | June 16, 2009 10:51 AM
Singularity Hub just published an opinion on the Lyons vs Kurzweil drama. You might want to check it out:
http://singularityhub.com/2009/06/16/kurzweil-in-the-lyons-den/
Posted by: Rev. BigDumbChimp | June 16, 2009 11:07 AM
Rev. BigDumbChimp loves when a blog goes all third blog on us.
Posted by: Knockgoats | June 16, 2009 11:27 AM
James Redford,
I don't have the maths and physics to assess Tipler's paper. However, I can read his final sentences:
"There
is some life remaining in the ‘old’ SM and quantum gravity based on the metric of spacetime
(and perhaps some life even in the far future of the universe). In fact, these theories may
be able to explain ALL observations, even the dark matter and dark energy, and why there
is more matter than antimatter. We may have already discovered the long sought Theory of
Everything!"
He seems rather less certain than you. I can also note that according to Google Scholar, this paper has been referenced by very few others since its publication. I would rather expect a successful ToE published several years ago - or even one which other physicists thought was worth taking seriously - to be the subject of numerous references from later work. Can you explain this anomaly?
Posted by: a_ray_in_dilbert_space
|
June 16, 2009 11:38 AM
James Redford,
Sir, you do not have the foggiest notion of what you are talking about when it comes to physics. There still exists no satisfactory quantum theory of gravity. We still have no theory of everything. And even the Higgs mechanism remains to be verified (though it's probably true).
The reason String Theory remains attractive is because it is plausible--as in nothing to date contradicts it. Same with Quantum Loop Gravity. It's plausible. Neither theory is verified.
Where on Earth are you getting this stuff, 'cause it sure ain't the peer-reviewed physics literature!
Posted by: SC, OM | June 16, 2009 11:59 AM
I think I read something online a few years back about some groundbreaking discoveries by a Dr. Pildózola Tuba, president of the prestigious Andean Physics Society...
Posted by: a_ray_in_dilbert_space
|
June 16, 2009 12:10 PM
Knockgoats, Tipler left the real axis for good in the 1990s. He went all Strong Anthropic on us. His work since then has been very speculative and not particularly useful.
Also, Tipler's whole cosmology requires the Universe to end in a big crunch singularity. The fact that the expansion of the Universe suggests in fact that the Universe is open would seem to rule that out empirically.
BTW, here's a theoretical physicist's take on Tipler:
http://www.nature.com/nature/journal/v371/n6493/pdf/371115a0.pdf
Tipler also features prominently Shermer's book: Why Smart People Believe Weird Things.
Posted by: a_ray_in_dilbert_space
|
June 16, 2009 12:16 PM
SC said, "Dr. Pildózola Tuba, president of the prestigious Andean Physics Society... "
Yeah, must be that--and the petrified flying saucers that some folks mistook for clams...
Posted by: Knockgoats | June 16, 2009 12:25 PM
Anyone followed James Redford's link? I just did. He thinks Jesus was an anarcho-capitalist.
*Backs away slowly, trying not to alarm the loony*
SC,OM
Nice one!
Posted by: Knockgoats | June 16, 2009 12:30 PM
a_ray_in_dilbert_space,
Thanks - I'll check that out.
Posted by: a_ray_in_dilbert_space
|
June 16, 2009 1:06 PM
More on Tipler:
http://blogs.discovermagazine.com/cosmicvariance/2009/01/05/the-varieties-of-crackpot-experience/
He's drifted further and further into the weeds until just about every nutjob is quoting him.
BTW, he's a good physicist. He wrote the texts I studied as an undergrad. Physicists are human, believe it or not.
Posted by: attoman | August 11, 2009 1:37 AM
Kurzweil is reputed to have produced a series of key basic inventions. I have been unable to find these key inventions from the 70's, 80's or later. He has some few patents in the 90's and this decade but they are hardly core or basic. In fact he seems to be attracted to the science fiction of Eric Drexler and his merry band of nanobrains.
It's kind of sad that he is unable to actually invent anything new that instead he is a master of hype, PR and pedantry.
One good piece of news the nanobots he predicted are here now! Not yet tamed of course the other futurists call them viruses but they are actually self replicating nanomachines that can travel through our bodies at will and into every cell, mitochrondria and gene in our body.
Watch out Ray the Swine Flu is not kosher!
Posted by: attoman | August 11, 2009 1:42 AM
If as was stated elsewhere here, IBM took down Blue before an independent scientific team replicated their demonstration at least once and preferably three or four times then the event is a fake lacking normal scientific proof.
A chess master must now be beat again(and scientifically tested) or the machine is still inferior.
IBM knows this and the rational of the world should know it too.