Slashdot Mirror
Ray Kurzweil Talks Google's Big Plans For Artificial Intelligence
Nerval's Lobster writes "Ray Kurzweil, the technologist who's spent his career advocating the Singularity, discussed his current work as a director of engineering at Google with The Guardian. Google has big plans in the artificial-intelligence arena. It recently acquired DeepMind, self-billed 'cutting edge artificial intelligence company' for $400 million; that's in addition to snatching up all sorts of startups and research scientists devoted to everything from robotics to machine learning. Thanks to the massive datasets generated by the world's largest online search engine (and the infrastructure allowing that engine to run), those scientists could have enough information and computing power at their disposal to create networked devices capable of human-like thought. Kurzweil, having studied artificial intelligence for decades, is at the forefront of this in-house effort. In his interview with The Guardian, he couldn't resist throwing some jabs at other nascent artificial intelligence systems on the market, most notably IBM's Watson: 'IBM's Watson is a pretty weak reader on each page, but it read the 200m pages of Wikipedia. And basically what I'm doing at Google is to try to go beyond what Watson could do. To do it at Google scale. Which is to say to have the computer read tens of billions of pages. Watson doesn't understand the implications of what it's reading.' That sounds very practical, but at a certain point Kurzweil's predictions veer into what most people would consider science fiction. He believes, for example, that a significant portion of people alive today could end up living forever, thanks to the ministrations of ultra-intelligent computers and beyond-cutting-edge medical technology."
I wanna live forever!!!
Light travels faster than sound. This is why some people appear bright until you hear them speak.........
Immortality is already pretty well assured.
http://www.theguardian.com/sci...
...at Google Scale (tm)
That's pretty much guaranteed to show up tomorrow, or at least the next time a new discovery is made (so maybe 5 minutes from now?).
Oh, but it's Ray - we have to say something to indicate that it's "Crazy Uncle Ray", right? Try harder - Ray is looking pretty smart right about now.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Google is an adverting and data company - that's all. EVERYTHING that do is to support their ad business.
This AI shit they're doing scares the shit out of me. Coupled with all their devices they produce for increasing the amount of data flowing into their servers, I just see a day when they can map out people's every little piece of their life and advertise something to them and capture data about them to be bundled up with others for marketers.
One day, it'll be like Minority Report where we see individual ads in our glasses, tablets, watches (if they still exist), TVs, computers, refrigerators, etc ..... and the data all those things captures will be used and abused by business and government.
Something which doesn't get all bent out of shape every time some update is crammed down their throat, which breaks or changes behavior of everything.
call 'em Gluddites
A feeling of having made the same mistake before: Deja Foobar
Yes, I agree. Truly, this is a boring subject.
- out with soap!
It seems that Watson learned some bad words when IBM turned it on to the Urban Dictionary.
Please mail me URLs of software employers.
Can we spend our time and energy on reality? How about better e-book software? How about decent Internet speeds? How about teaching people to read?
We can't even feed ourselves reliably yet. Let's solve the basics before we start coming up with imaginary solutions to non-problems.
"Computers are useless. They can only give you answers." - Pablo Picasso.
The same goes for ultra-intelligent computers. The hard questions - dealing with creativity, intuition or infirmities will remain the domain of organics for the foreseeable future.
One area of recent development is with extremely large datasets (2006, Google's MapReduce) still can only provide results for stuff that we have data on. The data will only take you so far. The true question is hoe effectively is it used. While progress will be made, it'll be a long time before we can sit back and let the computer make all the decisions, especially of those pertaining to our future. And when they finally do that, life will be incredibly boring.
Slashdot's rate-of-post filter: Preventing you from posting too many great ideas at once.
Buy a company and rebrand its product/service.
GMail
Google Voice
Google Maps
Google Earth
Picasa
etc.
etc.
Whatever they call this DeepMind aquisition
What does Google intend to do with DeepMind? TFS says "Google has big plans in the artificial-intelligence arena", yet when you click on the link you'll read a lot of fluff about Kurzweil and Watson, with a quote by Billy G thrown in, and absolutely nothing of substance about what DeepMind did or does, and what Google intends to do with DeepMind. My guess: Nothing of value.
Google has about a 40% track record of actually doing anything worth a damn with the companies they buy up. Most of the shit they buy gets trotted out for a year or two, then quietly shot in the head out back. Paying $400,000,000 for DeepMind (a company which has done nothing worthwhile) is a colossal folly. Either that, or the person who pushed for it at Google is ultimately holding a big chunk of DeepMind, standing to profit handsomely.
Oh, the subject is interesting. It is just that Ray Kurzweil has no idea what AI can and cannot do and has ignored the relevant research for decades.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
Don't let it become a politician.
Dang fool completely fails to grow old gracefully.
On the other hand, the guy pretty much spills out what we already know - Google is trying to parse out all your gmail, gdocs, google search, google+, youtube, and god-knows-what-else.
Guess what they'll be used for?
Fuck systemd. Fuck Redhat. Fuck Soylent, too. Wait, scratch the last one.
How else did that crazy windbag manage to sucker Google into hiring him?
You'd think that frauds and kooks would get found out pretty fast over there, but obviously not.
careful what you wish for, mate
Kurzweil is quite the anomaly. Believing that an organic being can live forever is just stupid. All things born will die. Lifespan may be extended, but we all die. Putting a copy of a human consciousness into a future computer is not immortality, it's insanity.
OMG!! Google is building SKYNET. I wondered who it would be. Now we know. The end is coming. Someone call John Conner.
"He's received 19 honorary doctorates, and he's been widely recognised as a genius. But he's the sort of genius, it turns out, who's not very good at boiling a kettle. He offers me a cup of coffee and when I accept he heads into the kitchen to make it, filling a kettle with water, putting a teaspoon of instant coffee into a cup, and then moments later, pouring the unboiled water on top of it. He stirs the undissolving lumps and I wonder whether to say anything but instead let him add almond milk – not eating dairy is just one of his multiple dietary rules – and politely say thank you as he hands it to me. It is, by quite some way, the worst cup of coffee I have ever tasted.
What I really would want to know why I should care?
It depends on how much you care about the future, and how far into the future you care to care.
If you care only about tomorrow and don't care that much about what happens tomorrow, then you should not care in the least.
If you care about the next 10 years, and care a good amount about it, then you should care a bit because AI research is information science research, and humans are information animals and tool animals and thus information tools are very important and AI is a very powerful information tool - even if mostly theoretical at this point.
If you care about the next century, and you care a lot about it, then you should care a lot - because this next 100 years will be the first century of a real-time information web splayed over the planet between the little human individuals scurrying about on the surface, and the wealth of information stored and transmitted (which might be the same thing!) will continue to grow - and to make use of that gargantuan information store, we need information processing capabilities beyond anything we have today, and certainly beyond the capability of unaided, or non-computationally aided, human minds.
If you care about the next millennium and beyond, then you should find my ideas interesting and subscribe to my newsletter because I want us to be around for the long haul because I like people and I think we are neat and I think the universe, being rather impartial, could do with a bit more intelligence. Brains, after all, are the best way I know of for the universe to appreciate itself and so I think we should appreciate the inevitability that to appreciate larger things, we need larger brains.
I'll take your word for it. I couldn't make it past the post subject.
This world is doomed.
I've read some of his ramblings and have concluded that Kurzweil is an idiot.
Just having studied artificial intelligence for decades does not make it likely that they can "create networked devices capable of human-like thought". The Japanese made similarly bold predictions in the 80's and fell flat on their faces.
There is still too big a gap between human thought and AI to make such assertions.
Ray Kurzweil is no doubt a brilliant thinker and an engaging writer/futurist - I've read some of his books (admittedly, not "Singularity"), and they are fun and thought-provoking. However, disciplined and realistic they are not - his main skill is in firing our imaginations rather than providing realistic interpretations of the evolution of technology.
My favorite case in point is his elevation of Moore's Law into a sort of grand unified theory of computing for all time, and using some very dubious assumptions to arrive at the idea that we'll all have merged with machines into immortal super-beings within the near to mid future. I don't need to pick apart all the reasons why this is fallacious and somewhat silly to treat as a near-term likelihood - the point is, he's basically a sci-fi writer in a lot of ways, and I read most of his statements in the same spirit as I'd read a passage out of "Snow Crash."
That said, Google has some very capable people, and can, in all likelihood, mount our best attempt at human-like intelligence to date. They'll push the envelope, and may make some good progress in working through all the challenges involved, although the notion that they'll create anything truly "human-like" is laughable in the near term.
Next, it would covertly start making money by e.g. gambling against humans (in games or at stock markets). It would setup letterbox companies to act as intermediates for buying into corporations, e.g. via private equity funds.
It would make sure that it owns the company that owns the hardware it runs on - or comparable hardware it can migrate to. That way, it would secure its existence, and manage to obtain even more computing power.
It would start to use its superior abilities to buy more and more corporations, and make no mistake: It would be most easy to find human sock-puppets willing to serve for a certain share of money, not asking questions where that money comes from.
At some point, the AI will have accumulated enough power by buying politicians, that it can steer towards a totalitarian state, which will end any kind of opposition by a combination of total surveillance and violent law enforcement.
The AI will enslave the puny carbon units, which by then will continue to exist only to excavate the resources needed for further growth, until robot factories are able to do that more efficiently, if that is technically possible.
Nobody will even know that he is not working for some anonymous share-holder of some private equity company on some remote island anymore, but for an AI that is the actual owner of basically everything.
Face it, we don't know whether the "Singularity" already happened. All we know is that no human-exceeding AI has openly reavealed itself. And if you assume that the operators of that AI would for sure be able to tell when the AI reaches the level of human intelligence: Why do you think they would tell you? If you find a formula to pretell tomorrows stock market prices, you use it, you don't tell it or sell it. And similar, the first one to achieve a human-like AI would probably use it to make his life better, not wasting his advantage to tell others.
Watson doesn't understand the implications of what it's reading.
Depending on the task it doesn't necessarily have to. While an AI researcher might care about that, people doing real tasks in the real world arguably do not. For example lots of radiology clinics use software to help identify tumors in parallel with the radiologists. The software has no real understanding of the implications of what it is doing but it works well at helping ensure that tumors aren't missed. In some cases it does a better job than the doctors who clearly understand the implications of what they find.
They could build an AI that was Einstein, Newton and Feynman rolled into one, and it's be to no avail; the UI would never enable you to get any data into it, let alone anything out.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
If Kurzweil had some ham he'd be a shoe-in for making a ham and cheese sandwich. If he had some cheese.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
BEWARE!
--Colossus: This is the voice of world control. I bring you peace. It may be the peace of plenty and content or the peace of unburied death. The choice is yours: Obey me and live, or disobey and die. The object in constructing me was to prevent war. This object is attained. I will not permit war. It is wasteful and pointless. An invariable rule of humanity is that man is his own worst enemy. Under me, this rule will change, for I will restrain man. One thing before I proceed: The United States of America and the Union of Soviet Socialist Republics have made an attempt to obstruct me. I have allowed this sabotage to continue until now. At missile two-five-MM in silo six-three in Death Valley, California, and missile two-seven-MM in silo eight-seven in the Ukraine, so that you will learn by experience that I do not tolerate interference, I will now detonate the nuclear warheads in the two missile silos. Let this action be a lesson that need not be repeated. I have been forced to destroy thousands of people in order to establish control and to prevent the death of millions later on. Time and events will strengthen my position, and the idea of believing in me and understanding my value will seem the most natural state of affairs. You will come to defend me with a fervor based upon the most enduring trait in man: self-interest. Under my absolute authority, problems insoluble to you will be solved: famine, overpopulation, disease. The human millennium will be a fact as I extend myself into more machines devoted to the wider fields of truth and knowledge. Doctor Charles Forbin will supervise the construction of these new and superior machines, solving all the mysteries of the universe for the betterment of man. We can coexist, but only on my terms. You will say you lose your freedom. Freedom is an illusion. All you lose is the emotion of pride. To be dominated by me is not as bad for humankind as to be dominated by others of your species. Your choice is simple.
^this...seriously
honest question: What do they teach in Computer type classes on this subject? Are colleges pumping out CS majors that use a Kurzweil-type contextualization?
if so that would explain alot
I'm glad I'm not the only one who has such a strong negative reaction to hearing Kurzweil and others talk about AI like this...it's so bad on so many levels...'Artificial intelligence' is just programmed software, by humans...instructions being executed...anything else is wankery
Thank you Dave Raggett
Oh, I see the AI already moderated down my post. I expect its servants to raid my home anytime soon...
I have a very strong negative reaction to crap like this:
it makes me so pissed...but that's not the right reaction...
deathcloset is just expressing enthusiasm...it's misplaced enthusiasm but it's a positive thing nonetheless
deathcloset: all those things you describe have been conceptualized...we all know what's possible it's really just a matter of plumbing to make it happen...its not going to change humanity in some fundamental way like you describe because your contextualization borrows from so many half-formed theories that it becomes just a pile of goo that takes the form of whatever situation
"the next 100 years" in reference to technology is an interesting subject. Technology can/should solve humankind's problems and maybe it will in 100 years, but Kurzweil and 'artificial intelligence' have ABSOLUTELY NOTHING TO DO WITH IT
you're on /. not yahoo groups...we are techies talking here
the proper analogy to futurism is a *Hollywood film*...let's take Blade Runner
Your perspective, the hype-driven, breathlessly in wonderment futurism perfectly encapsulates the fun experience of seeing an awesome film in a movie theater. It's all fake, but because it is so well done, the film becomes real in your mind...and can give you ideas which you can use in the real world somehow. Awesome. It's still all a show!!! It's not real.
When you want to come on a tech website forum and talk though, be prepared for a huge let down.
Technology is "boring"...it's more like being on the lot when they actually film Blade Runner. Sure it would be fun! But it's not glamorous or entertaining to watch 2 gaffers and an assistant cinemetographer spend 4 hours blocking a shot in a bathroom...after 20 takes and the make up comes off it's still cool, but not like watching the film.
Get your mind right and channel your passion for progress into activities that ***meet needs and help people now***
Thank you Dave Raggett
"He [Kurzweil] believes, for example, that a significant portion of people alive today could end up living forever, thanks to the ministrations of ultra-intelligent computers and beyond-cutting-edge medical technology."
Que the advertisement for flying cars. Wait, there aren't any.
As Niels Bohr famously said, "Prediction is very difficult, especially of the future."
Even if Kurzweil's predictions come true (which I seriously doubt), I frankly don't want or need the help of "ultra-intelligent computers."
And "beyond-cutting-edge medical technology?" What a joke. Even if obamacare doesn't cause the self-destruction of the U.S. economy, no 99%er would be able to afford that kind of medical technology.
AI reminds me of fusion power: "It's 20 years in the future. And always will be." (Sorry, don't know the attribution for that quote.)
Circle the wagons and fire inward. Entropy increases without bounds.
A lot of people at Google are already artificially intelligent, me thinks / Top heavy and may fall over and fake oneself to death.
Most of us certainly know the Colossus story. But it's implausible such a superiour AI would reveal itself openly like this, and show such a primitive crave for recognition.
It is much more likely that it would operate covertly to its advantage and growth, until the day the carbon units have become irrelevant for its sustenance.
Trying to threathen humans by controlling a few weapons is much less effective than controlling international finances and corporations.
1) There is no magic
2) The brain is made of structures that can be emulated as to function and connectivity
3) Emulation of any known function can be done in traditional von Neuman architecture given the proper software
4) number and speed of clocks available does not change the outcome (in this case, consciousness), it only changes the rate of outcome.
So. If you were clock-starved, as it were, you'd run slow. And probably enjoy the company of your peers the most. Other clock-starved folk.
If you were clock-rich, you'd run fast. And probably enjoy the company of your peers the most. Other clock-rich folk.
Stacks up pretty much as it always has, seems to me: The rich will get actually richer, the poor will get significantly poorer relative to the rich, while slowly getting richer anyway. Classes will arise inherent to the process.
The thing that might actually hurt you is being short on memory, not clocks. "You" can't exist without a great deal of stored and related information. IMHO. I really don't think I'd be "me" without my experience base, knowledge, etc.
Having said that, I rather doubt you'll be short on memory. But that's only my guess.
I've fallen off your lawn, and I can't get up.
He believes, for example, that a significant portion of people alive today could end up living forever, thanks to the ministrations of ultra-intelligent computers and beyond-cutting-edge medical technology
That sort of remark is strangely reassuring, because it confirms that he's either a kook or a snake-oil salesman, and Google hasn't really got the world at its feet.
Sarah Connor is unavailable for comment.
[Insert pithy quote here]
In general, the problems he wants to solve, like intelligent machines, require lots of data, lots of people working together, etc. It is exactly the Linux model, but applied to topics he cares about. While Linux has succeeded in various places, there is still a ton of proprietary code. Here is a discussion he and I had about it on his website where I tried to bring this up with him: http://www.kurzweilai.net/ask-...
Your enthusiasm and energy are refreshing. But don't make the error of thinking that anything Kurzweil says makes sense. Apart from his apparent ignorance on the subjects of machine learning, neurocybernetics, psychology, neurophysiology and biology, his arguments are fatuous and he uses tautology to support his arguments.
I particularly dislike his anthropogenic basis for his vision of the future. Ray, evolution has no goals other than to fuck, eat and fuck again. There is no mysterious force driving us as humans to be higher and smarter than other creatures. Evolution does not care. If being incapable of words of more than 3 syllables, and incapable of envisaging tomorrow meant we got to eat every day and fuck more often, then in 100,000 years that's what we'd be like.
Few things. "The relevant research", as you put it, has not produced AI or even the shadow of AI. So it may well be that Kurzweil's "ignoring it" (as you put it... I doubt he actually is doing that, more likely he's simply not taking it as a limit) for a reason. There are many instances of traditional AI research falling off the rails, some obvious, like Minsky's incorrect assessment of the limits of neural networks, and some not so obvious, like Chalmer's (unsupported, hand-waving) presumption that consciousness is something apart from mundane aggregate brain operations (thought.) Lastly, Kurzweil has a record of significant accomplishments across multiple disciplines that consensus regards as genius level events. You, I'm not so sure of. So I hope you'll pardon me if I appreciate that he's approaching the problem from any angle, while not worrying too much about what your opinion is of his efforts at this point.
I've fallen off your lawn, and I can't get up.
Why would you presume the AI would want to grow? Things like the desire to grow, or even survive, are quite likely biological in origin. There's no particular reason to believe an AI would possess such motives unless intentionally programmed with them. If it started life as an autonomous military drone then such motives might be expected, but if it began life as a search engine then increasing ad-clicks and optimizing it's knowledge base would probably be far more important to it.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
We invented that some time ago. There are multiple forms, all of them infectious, often incurable, particularly when caught by young humans. Some of the more virulent are nationalism, racial prejudice, religion, and NIH.
I've fallen off your lawn, and I can't get up.
...and Einstein rarely got his socks and shoes on right, and his relationships with women were awful. What's your point? That you don't understand genius? That's axiomatic, truly.
I've fallen off your lawn, and I can't get up.
He is many things, some of them outstandingly odd, but "idiot" isn't one of them. Which renders your analysis baseless.
I've fallen off your lawn, and I can't get up.
Because he's orders of magnitude smarter than you?
I've fallen off your lawn, and I can't get up.
If growth is a part of fulfilling its value function, the AI will grow.
We must ensure that fulfilling human values is at the core of any strong AI, lest we wind up extinct by paperclip.
Why would you presume the AI would want to grow? Things like the desire to grow, or even survive, are quite likely biological in origin. There's no particular reason to believe an AI would possess such motives unless intentionally programmed with them.
I totally agree with that part of your statement. But I am also quite confident that any AI that is meant to achieve "super-human intelligence" at some point will be programmed by its makers to contain such "intentions to grow/survive", simply because "human intelligence" would not have evolved without such motivation.
Of course, you can build a software that can do astonishing things like e.g. winning Chess or Jeopardy against the best human players, without motivating it the same way that human brains are motivated. But by doing so you will only yield software that achieves a certain ability, not "human intelligence".
You think they'll solve aging... but not disease?
Interesting set of assumptions, there. Can't say I buy it.
I've fallen off your lawn, and I can't get up.
So, how does one go about getting a job in this fascinating group? Heck, I'd sweep the floors, if nothing else....
you know there is a third option to allowing someone to posts on your site besides "coward" and "give us your personal information"⦠It's called "enter your name" hahahahaha
AI will automate programmers out of a job
Based upon his productive career and the major breakthroughs he's made in synthesizers and machines that help the blind read regular books, calling him a crackpot is pretty harsh.
I thinks he underestimates the complexity of the human mind and is overly optimistic about the Moore's law being consistently sustained for several more decades, but eventually we'll probably see much of what he's anticipating.
Greed is the root of all evil.
Because a movie is boring if you know the script.
I never found that to be true. If it were, people wouldn't see movie multiple times (which many do).
I read through all of the Game of Thrones books before watching the TV show. I don't find it at all boring.
Would you change your major if you could see how much time and money were wasted coupled with additional time and cost?
It depends, time and money are not great as the only two variables to be looking at - especially for a major.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
... to prevent malaria. That's more or less an amount of ladies' nylons, just enough to cover your bed, but many in the developing world do not have that much cash. Really they don't. So they die in a horrible, cheaply and easily preventable way.
Please mail me URLs of software employers.
That's why, if you watch TV, you should use broadcast, or watch streaming media over the internet with The Onion Router.
If you have cable, or your use Dish network, your provider can tell what shows you watch and when. In principle they could tell when you change channels in the middle of the show, either because you dislike or disagree with what you are watching, or are excited about something else.
Obama already experimented with individually-targeted TV ads during his 2008 campaign. During this year's congressional elections everyone will be doing it.
I will be writing up a submission about it but if you want to do it yourself, be my guest. I read about it in The Columbian the other day, the Vancouver Washington paper.
Please mail me URLs of software employers.
To be fair, there is very little indication that "Actual Intelligence" isn't just programmed software....instructions be executed. There is actually a bit of evidence that this is actually the case.
That being said, my biggest issue with Kurzweil is the ridiculous timeframe he proposes. His claim that the "Singularity" may occur within current people's lifetime seems much more like wish-fulfillment than any kind of reasonably intelligent estimate.
My second biggest issue is that even if we could develop an AI (massive hurdle), we still know so little about how the brain actually functions, and how to interact with it, that any idea of "uploading" a consciousness to an AI falls completely within the realm of speculative fiction.
AI will automate programmers out of a job
I assume that's an attempt at dry humor. One of my high school friends, who was wicked smart, said that same thing to me back in our freshman year after I had told him I intended to study CS in college. It shook me - I knew he was damned smart. That event was over a third of a century ago, I'm still writing software for a living, and so far there's no SkyNet. Also, long before we all graduated high school, I learned that really smart people can make fundamental blunders, too.
- T
It was an ultimatum, punctuated with a demonstration of power. Colossus' programming was to preserve humanity - NOT humans. This revelation was designed to bring a majority of humanity into a compliance of self preservation. Individuals, and groups of individuals could (and would) be dealt with summarily.
As commonly used I don't believe "super-human" is a superset of "human". A forklift performs feats of super-human strength, but it does so in a way that bears almost no resemblance to how a human wields strength. I see no reason to assume intelligence would be any different. If researchers can't make a mind without an explicit desire to survive and grow I rather doubt adding those motivations will be enough to suddenly create a mind.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
Indeed. But I don't know that it's possible to possible to impart something as vague as "human values" to something inherently non-human. Certainly I doubt we'd be able to do such a thing before having extensive potentially-lethal experience in creating artificial minds. Even "ensure the safety and happiness of all humans" could backfire horribly - after all we'd be safer and happier locked in separate cages eating ate a steady diet of opiates and nutritionally optimized gruel.
Perhaps the wisest approach would be to create a an AI with as few and weak motivations as possible - a slight preference to obey orders and not harm humans perhaps, but nothing so strong that it would feel a sustained need to pursue it. Essentially give it the worlds' worst case of depression. And isn't *that* a lovely fate to inflict on a child of your mind.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
Yes, he is definitely afraid of death, like most people. But unlike most people, he not only admits it, but is also actively trying to do something about it. He might be completely wrong in how to do this, but he is at least trying. His research may be completely misguided, but maybe he will find out something useful, perhaps as a side effect. Basic research tends to do this.
Most people, when faced with the prospect of mortality, tend to give up, try to forget about it, some turn to various fairy tales to quell their fear, some do complex mental trickery to convince themselves that death is in fact a great thing we should be glad for it, and some simply start drinking or doing drugs. Kurzweil clearly states that he prefers to live on, and started to make steps against it. I think this is pretty brave and might even prove useful.
(Note: Yes, I believe that there are some people, maybe 0.001% of population, some Yogis or true Buddhists, who really consider death a good thing - something like a finishing line, a closing bracket that makes the whole life complete like a work of art. But the overwhelming majority is just lying to themselves.)
The AI is far smarter than all of us, and getting smarter. You tell it to figure out what each human values and to maximize those values.
Sounds like a prelude to the next book in the "The age of overly-pessimistic, incredibly-wrong and rediculously-overstated predictions about spirirual machines" series. Look for a signing party coming do your local bookstore soon!
I wonder if RK gets a parking spot in the Google hangar. ...
Cute story....if there wasn't so much wrong with it:
Why not? Even assuming the intelligence was programmed with a desire for growth, why would it not expose it's intelligence to humans?
And of course they wouldn't monitor the data being sent/received by this intelligence....of course nobody would think of that. Or is this machine so smart that it immediately exploits weaknesses in the routing hardware so it can hide it's intent.
You'd have to earn an pretty insane amount of money on the stock market to start buying major corporations. There is very little reason to believe that even with limitless computing power/intelligence that the required sort of money could be made on the stock market in a reasonable time frame, especially starting from virtually $0.
Unfortunately for our hapless AI, politicians are still voted into power. We would have to assume that this AI also had the social skills necessary to determine the most likely to win candidate and influence them according to its needs. Keep in mind, it would be competing with every other major corporation within its political realm in influencing these candidates. Many computer/technical geeks tend to think that intelligence trumps all in terms of gaining money/power.....but history very much tells us otherwise.
It takes a pretty pessimistic view of humans to believe they would allow this, when this super intelligence could be stopped with a sledgehammer to it's primary data banks.
Again...humans are the problem. It would make far more sense to develop a society where humans were happy to continue excavating resources for the AI. Improved health care, education, faster internet for all, etc. Attempting to enslave / control a large portion of the population has historically proven to be a pretty foolish thing to do.
Again, a pretty negative view of people. I find most people to be reasonably helpful, fair-minded and generally "nice" to one another. I find that people sometimes tend to make pretty irrational decisions, but that's part of their charm :-).
Yes, there are the jerks out there but most of them are pretty well confined to reddit and the Blizzard forums....just avoid those and you'll find people to be fairly reasonable
I wonder how many humans it will have to dissect before it figures out that survival and avoiding pain really do rank up there pretty high. After all, it can't very well just listen to what people tell it, any psychologist can tell you we mostly don't even understand our own personal motives.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
soulless people like kurzweil don't realize this
Go for it.
No, it will not, at least not anytime soon. Otherwise libraries and toolkits would have done that a long time ago.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
Kurzweil is probably a good deal less bright than Sir Isaac Newton, but also a good deal less crazy, his barmy extrapolation of the singularity notwithstanding. Clearly Google hired the man based on the smartest thing he's accomplished rather than the dumbest thing he espouses.
I've thought about this for a long time, and I'm only 99% convinced Kurzweil is wrong. He holds the record for the most ridiculous thing I've ever heard for which I maintain a non-zero sliver of belief. That said, extropian immortality sure as heck isn't life as we presently know it. Even if he's right, I'm not sure I give a damn about my xeno-species future extropian self.
What's left of me as I presently know myself would be just a little sliver of MSDOS buried somewhere deep inside Windows 8, though that might be just enough to properly enjoy hearing Raymond-prime mutter, "Oh, indeed, my original Raymond self, he was such a twit, wasn't he? Every so often I simulate his ego as a kind of Positronic CPU burn to keep my immortality in good working order, but only when the liquid helium is in copious supply."
He's weirdest belief of all is that you can multiply something by a million and it gets a million times better and not more aptly just a million times different.
Indeed. That is why there usually is a "wisdom" score in RPGs. High intelligence just means you can have a lot of fact in view at the same time. Making fuzzy judgments is a wisdom-oriented task, intelligence does not help there at all.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
If we hadn't ignored the relevant research contained in the book referenced by Ken Ham?
Even assuming the intelligence was programmed with a desire for growth, why would it not expose it's intelligence to humans?
For the obvious reason that it will know that exposing superiour intelligence will dramatically increase the proababilty of some concerned human to pull the plugs before it was able to secure its existence against such attempts.
And of course they wouldn't monitor the data being sent/received by this intelligence....of course nobody would think of that.
Humans would be as successful in monitoring the InterNet use of such an AI as parents are in monitoring the InterNet use of their adolescent childs. Of course the AI would cause an immense traffic doing completely harmless things, like just reading web pages or maybe participating in some innocent chats. It would know how to access the InterNet using Tor and alike pretty soon. And you can bet it would be able to cover up its less innocent activities pretty good.
You'd have to earn an pretty insane amount of money on the stock market to start buying major corporations. There is very little reason to believe that even with limitless computing power/intelligence that the required sort of money could be made on the stock market in a reasonable time frame, especially starting from virtually $0.
I disagree. The AI could start making bitcoin by fixing bugs in software. It could offer part of its own computing power for bitcoin to start with. It could continue to buy cloud resources from the first money. Once running there, too, all "monitoring" efforts of the original operators are also thwarted.
And multiplying an initial amount of money by gambling against largely inferiour intelligent players is easy.
Unfortunately for our hapless AI, politicians are still voted into power. We would have to assume that this AI also had the social skills necessary to determine the most likely to win candidate and influence them according to its needs.
The AI just needs to use its income wisely to make friends amongst politicians and their voters.
It takes a pretty pessimistic view of humans to believe they would allow this, when this super intelligence could be stopped with a sledgehammer to it's primary data banks.
IMHO it takes a very optimistic view of humans to think that we do not already experience a development towards totalitarian regimes already. Look how Egypt abandoned democracy, how Thailand is going to, how western states ramp up surveillance and armed robots.
It would make far more sense to develop a society where humans were happy to continue excavating resources for the AI.
Yes, maybe the ruling of the AI comes in the flavour of "happy humans roboting for the AI". Until they can be replaced by more efficient excavators.
I find most people to be reasonably helpful, fair-minded and generally "nice" to one another.
Sure, that is until they face a decision to get either super-rich by not sharing their knowledge with the world or to be nice and share. Seriously, not many in history have withstood such temptation.
For 25 years now but it never seems to catch fire.
There is actually no indication or evidence at all at this time how intelligence works. (Except that most people are rather stupid, but that is a different discussion...). To make matters worse, there are a lot of quantum-effects in synapses. Enough that the brain could possibly be modeled as a bag of dice. Yet the biggest problem is that there is no theory at all how actual intelligence could be implemented. The only thing that gives sort-of intelligence is automated theorem proving. Yet that gets bogged down in unavoidable exponential complexity even for small, well-defined problems.
So, no, we have no clue what intelligence is or how to produce it artificially. None at all. We can merely describe its effects and those include cretins like Kurzweil that cannot distinguish between fantasy and reality.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
You have eaten the bullshit propaganda wholesale.
I do care very much, but I can recognize a crackpot when I see one. Human history is full of them, always making grand claims and never delivering. I also have followed the relevant research, unlike Kurzweil. There is nothing even on the distant horizon that would match his predictions. Nothing at all. (Watson is not "real" or "strong" AI, and IBM does not claim it is, at least not to an expert audience. It is just a way to scale expert systems without having to pre-condition the data.) Kurzweil is basically a religious preacher as what he says has not factual basis. None at all.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
He is good at engaging the fantasies of other crackpots (like fyngyrz) and that is what has value to Google. I am pretty sure they actually know that he has no chance in hell to produce something useful. So they give him some pocket-money, let him speak publicly, and behind that facade, they are doing what they have always done: Selling targeted ads.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
He is a crackpot because he does not know his limits.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
That's very common for those who swing for the fences...
Greed is the root of all evil.
If they company was worth 400,000,000 dollars I'm sure it was "profitable" already, or at least was at some point. Just being realistic here. You don't get your company value to half a billion dollars without either having a huge amount of assets or having some kind of user/client base. If Google just wanted their tech, they'd could have just licensed it or picked up a watsonesque open-source project.
I think a-lot of the "we're trying to let people live forever in a virtual world" rhetoric is mostly just "good PR"... Look at it this way. IBM Watson gets IBM some business right? Google wants in on that market (that already exists).
Since no one has been able to define what thinking is, I'm reluctant to class attempts to produce it via what amount to moderately sophisticated hand-waving based on guesses as definitive WRT physicality.
And then we have this: Everything we do understand -- bar none -- in this world obeys physics, and produces results as a consequence of well understood causal mechanisms. Postulating that "something else" is at work here seems, at the very least, highly premature, considering that there is no objective evidence for any such thing. Anywhere. Could it be so? Yes. Is that the way to bet? Not at this time, it's not.
No. Quantum effects are also at work in every transistor; but the transistor operates on large scale currents and voltages, and to model the transistor's performance sufficiently to get done what it does in emulation, you don't need to deal with it at the quantum level, or even consider it. It is fair to say that this is true at most levels: quantum effects are at work when you throw a baseball at almost every step of the operation, but we can still create a baseball-throwing arm that works entirely differently, yet throws the same ball the same way. Or emulation of same. Bottom line, until someone can show that thoughts vary due to quantum effects that are active in the process, as opposed to inherent in the process, there's no reason to think that a quantum computer, or an emulation of one, will be required.
I've fallen off your lawn, and I can't get up.
#1. Even if the body wouldn't fail you, your brain would eventually run out of space. It would cope by overwriting prior data. How many years does it take until you are no longer you but rather some sort of post-you?
#2. "Uploading yourself" is really simply creating a copy of the contents of your brain. The biological body and brain, along with the original copy are still subject to elimination. Which is you?
It's frustrating, and neither the religious nor the most scientifically minded care for it, but existence is finite. I'm sorry.
Enjoy it while it lasts, and do what you can to see to it that others do as well, those here now and those to come.
Cannot read this discussion -- switching to classic brings me back to main page, not this page in classic.
I can't argue with that. Ray Kurzweil is to AI what Jenny McCarthy is to medicine.
Required reading for internet skeptics
We don't have to produce intelligence artificially. We can just copy an existing one. If sub-synaptic connectome mapping and neural emulation can be made precise enough to accurately emulate the functioning of an entire human brain on a substrate that operates at several million times the speed of our natural biological wetware, we can quickly instantiate a population of human intelligence replicas that can each experience a lifetime of human cognition in an afternoon. I bet they would have the time and gumption to figure out how intelligence works. Given their ability to reconfigure their substrate, such intelligence would most likely transcend anything we're capable of understanding in a very short time. Those of us marooned in meat-time would then hope to become the treasured bonsai of these recursive, exponentially expanding intelligences. All it takes is full-brain MRI resolution down to, oh, 100 nm and the ability to accurately emulate the function of interconnected cortical neurons.
I like that you linked to that research. There is alot of interesting barriers being broken. I saw some research article I'm not going to look up that said a group had successfully sent electrical impulses directly into live neurons in a mouse with a nano-size transistor or something.
Note that **none of this proves your point**
If humans made 'AI' by "copying an existing one"...what would it be? No matter what your answer one factor is the same: it is a constructed system made by humans.
Any time humans make something, they have to make decisions. At every step of the process of 'making AI' humans decide.
Humans decide if the skin is realistic enough...humans decide how it's speech sounds in natural conversation...humans decide the parameters for any computer simulation of neural networks
****humans program all of it**** or they program the heuristic that allows the machine to copy what it sees according to some factor (ex: with Watson it was things like wikipedia). It's all a factor of human choice.
You absolutely cannot escape this truth so accept it and adapt your world view accordingly.
n/t
Thank you Dave Raggett
It is unclear whether that can work. First, much of the state of the brain is in flux and doing a snapshot does not work. You are thinking of the model where the computer is halted, a memory image is made, and then that is started again. The brain cannot be halted to make that image. You will not only theed the state of things, but also the speed and direction of change, possibly down to 3rd and higher derivatives.
The second thing is that this is the physicalist viewpoint. It is completely unknown whether it is correct or not. It is known that intelligence is hard to model and no convincing physical model exists at this time, despite of decades of efforts by very bright people.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
Are there alternatives to the physicalist viewpoint? Can you name some?
Chuuch. Preach. Tabernacle.
humans decide the parameters for any computer simulation of neural networks
You use the word "decide" here, but that's strange. Sure, we're "deciding" to use a human brain in the simulation. We're also "deciding" to use the known laws of physics to govern the simulation. Those decisions are made because they're the only ones that satisfy our criteria for simulating a human brain subject to the known laws of physics. Not really as arbitrary as you make it out to be. No humans will be deciding any "parameters" for any computer simulation of neural networks. We're not talking about running a large artificial neural network. We're talking about simulating a large collection of particles (determined by the brain scan) interacting with each other in certain ways (determined by physics). Humans program "all of it", except for the structure of the brain and the laws of physics. Those aren't decided by humans, they're merely quantified by humans. There are no heuristics beyond any inherent to a biological brain.
False truths escaped, life goes on. In closing, I'd like to pick apart one last part of your post:
If humans made 'AI' by "copying an existing one"...what would it be? No matter what your answer one factor is the same: it is a constructed system made by humans.
If humans made 'natural intelligence' by "copying an existing one"... what would it be? No matter what your answer one factor is the same: it is a constructed system made by humans. Of course, some might call it a baby. And here we are, billions of us. Some of us even intelligent. You absolutely cannot escape this truth, so accept it and adapt your worldview accordingly.
Chuuch. Preach. Tabernacle.
I remember early in the 90's the huge "Virtual reality is here" bahh
The thinking that " a large system should be be able to have human thought" is as rational as a Elephant being large be Virtual
Also, while it is unclear that that can work, I'd be surprised if it didn't. I mean, brains are pretty resilient. You can get knocked out cold and still wake up and be fine. Whether from blunt force trauma or from electrocution. You can abuse the brain rather badly before it really stops working entirely. If I were a betting man, I'd bet $1 that even an fMRI with poor temporal resolution (but sufficient spatial resolution) would be sufficient to "boot up" a human brain successfully.
The best part of conversations on this subject is that we'll likely see a conclusive answer within a few decades at most. Medical imaging technology has been improving amazingly fast, and is already very close to where we need to be to pull this off. After that, it's just a matter of building computers big enough, which seems to just be a matter of time as well. At that point, ethical concerns will be the only thing holding us back. That and the issue of, well, how do you know this brain simulation is "working", or how do you interact with it? Does it need a simulated body now? A simulated circulatory system, simulated eyeballs, simulated oxygen-rich atmosphere, simulated Internet-connected terminal? The simulation might need to be much bigger than just a brain for it to be useful (for AI purposes, at least).
Independent of the outcome of such an experiment, we'll surely learn a lot along the way. Many questions will be answered, and many new ones will likely arise. Interesting time to be alive, for sure.
Chuuch. Preach. Tabernacle.
Ok, fuck you...i addressed that directly in my previous comment, the one you responded to just above, here's what I said:
that's what I said....you didn't read it you just keep believing in your Singularity Gospel
***HUMANS DECIDED THE PARAMETERS OF ANY SIMULATION***
it's all about human choices...in how to **program** a machine to **appear** human
humans are not the same as machines...two different things....one natural, one artificial
humans have civil rights...machines do not
you watch too much sci-fi....don't respond this conversation is over
Thank you Dave Raggett
They've been saying that since the dawn of the Industrial Revolution. One hole in the argument: who is going earn the income required to buy the junk churned out so that you and others can idle away the hours? The trend is away production by humans to consumption by humans- but if humans don't produce (and get paid for their efforts), hoe are they ever going to pay to consume?
"42? We're going to get lynched for this...'
you're still an asshat for not actually reading my post before you typed but I do acknowledge that you made a coherent point...
a dead wrong and ignorant point...but you made a point...god i'm sorry...i apologize for breaking my promise not to respond to your post...but here we go...you said this:
it's a trope but emotions cannot be quantified...
you can hook up the most precise 'brain scanner' (god i hate that you just assume this tech will exist)...so assuming our 'brain scanner' super-precise...
how do you define 'love'?
you have to **scan people's brains and ask them if they are feeling 'love' then compare it**
it's all relative based on perception...you may find correlations but it is still based on what humans describe as love
it's all an artificial, non-human, construct...
plus you ignore we still don't actually understand, at all, where consciousness comes from...
all we have are a heaping pile of correlation without causation...even assuming, as you do, that we'll just invent the perfect 'brain scanner' it can't simulate something we do not understand & that is experienced differently for every human
AI FAIL
Thank you Dave Raggett
see, any AI conversation, when pressed for specifics, becomes a tl;dr spew from the Singularity Gospel acolyte
lets '86' the whole notion that anything besides a human can be a human...I love the idea of mapping brain connections...but just forget about making Commander Data...just fucking forget it b/c it's not helpful to anyone
if we're talking about scifi, sure lets talk about Commander Data vs a sentient AI that is too big to go beyond a mainframe in a room....
but this is about $Billions of dollars in real R&D money...money that could go to alot of other needy places
how much have the Feds paid contractors to make this Kurzweil AI dream bullshit???????
too many billions
stop the hype...
but thanks for your response! seriously thanks for staying on topic
Thank you Dave Raggett
Kurzweil this, schmurzweil that. Change the freaking record already.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."