Slashdot Mirror
Turing Test 2: A Sense of Humor
mhackarbie writes "Salon has a great story, Artificial Stupidity, about the Loebner Prize, a yearly contest that for over 10 years now has offered a $100,000 prize to anyone who can create a program to pass the Turing Test. The best part is the resulting fiasco that develops between the eccentric philanthropist who started the contest and extremely annoyed AI Researchers such as Marvin Minsky."
All Hugh Loebner wanted to do was become world famous, eliminate all human toil, and get laid a lot.
does this mean we're all considered entrepeneurs ?
Fear Breeds Knowledge
and extremely annoyed AI Researchers such as Marvin Minsky
This person is commonly known as Marvin The Martian.
Life is the leading cause of death in America.
I don't think bots are the problem... I've had several online conversations which I'd assumed were chat-bots but turned out to be real people. I guess when Turing designed his test, he probably didn't anticipate the massive advances in human stupidity that we've witnessed in the last few decades :)
-- Open Source: It's mad, but you don't have to work here to help.
Every person with whom I spoke about it said that last year's contest was an utter fiasco, with unclear rules, inconsistent judging, arbitrary fiats by an opaque prize committee, petulant prima donnas, and last-minute changes of venue that prevented most entrants from even discovering where the contest was taking place until after it had happened.
Was it held in Florida as well ? Or is it just a massive coincidence ?
Er, did you read the article? Funny thing about articles, they often have information very relevant to the corresponding Slashdot story! I know, it's hard to believe! I highly recommend reading one sometime.
Since 1989 Loebner has spent, by his account, more than $200,000 and a thousand hours of unpaid time to hasten the arrival of intelligent machines. He has set aside a gold medal and $100,000 in cash for the creator of the first machine that can pass for human. In the meantime he gives out annual prizes for programs that come closest to a long-sought holy grail in the artificial intelligence community: passing the Turing test.
slashdot!=valid HTML
IMHO, is the line that says: "for the past 25 years, AI specialists have been saying that all AI problems were going to be solved within 10 years" (or something like that).
This strikes me as true: for years and years and years, researchers have been promising AI was just around the corner... And what do we have right now? Nothing!
I want a Turing-class AI or my money back!! =)
The right to offend is far more important than the right not to be offended. (Rowan Atkinson)
All Hugh Loebner wanted to do was become world famous, eliminate all human toil, and get laid a lot.
Remind anyone of Hugh Hefner?
(overall a good read. certainly a buttload of speculation but no more (actually probably less) than found in Wolfram's book)
On the other hand, I see nothing wrong with offering a prize for what he believes in. Heck we have the Templeton prize out there (more than the Nobel, no less) for best achievement in religion (christianity specifically, methinks), so what's wrong with offering 100G of his own money? We also have the X-BOX cracking contest - who is willing to bet that the believing in the chance of solving a 2048bit key in a few monthes is MUCH dumber than trying to shoot for some "not everybody agree as AI" AI?
My life in the land of the rising sun.
Any specified Turing Test can be defeated in much the same way as a lock-pick can defeat any specified lock, so perhaps we should move up one level of abstraction. I propose the "Meta Turing Test" which is as follows: specifying the conditions of the Turing Test (ability to lie, sense of humour, etc.) should allow a true human to design an automaton that fools the turing test, while a computer will not be able to do so.
Alternatively, why not just abandon the myth that human intelligence is some kind of mystical cloud, and see it for what it is, namely a set of thinking organs each designed (or adapted, if you prefer the 'evolution is a passive process' concept) to solve specific problems, in the same way as my hand is adapted to handling objects. Then, test each of these tools carefully. Anything - computer or human - that passes the tests can be defined as 'human'. Many beings that we today consider human will probably fail. Borg borg.
Sig for sale or rent. One previous user. Inquire within.
Why are Minsky and Shiber so upset that a sex-addicted pothead is sponsoring an A.I. prize, when the Father of Dynamite sponsors a Peace prize?
Loebner can do whatever he wants with his dough. No one is being coerced into entering his contest.
I'm not really a web designer, I just play one on the Internet.
A. Rightmann
oh.
Malike Bamiyi wanted my assistance.
"... extremely annoyed AI Researchers ..."
....
Perhaps "extremely annoyed" is what distinguishes human intelligence from machine intelligence?
In John Brunner's non-novel Stand on Zanzibar, cranky sociologist Chad Mulligan declares that supercomputer Shalmaneser is now intelligent because Shalmaneser has displayed the quality of "bloody-mindedness". Not the same as "annoyance", of course, but in the same emotional realm
-kgj
I think that people who focus on the Turing test are missing the point, this isn't really AI and probably doesn't have much of a use outside advertising to via IRC/personal messaging etc.
The real interesting areas of research in AI are for example: in dye-master processes, where AI replaces a highly skilled human, or automating the driving of cars. These are all AI and, IMHO, much more impressive than glorified Eliza, Turing test stuff...
keyboard not found, press F1 to continue
Cheers,
Ian
Turing's 'imitation game' is now usually called 'the Turing test' for intelligence.
Hmmm. I'm pretty sure that there already are computers that would seem more intelligent than some of the people I've had talked to while playing CS.
Can I bum a sig?
the program that alters the test to fit its own capabilities. That is cheating? How more human can it get ? Humanity is constantly adapting it's surroundings to fit its own needs...
Actually, 100k is about the price of an undergraduate degree at a 4-year private college. Not sure that it would help w.r.t. the spelling though.
Physics of Consciousness
Building a machine to pass the Turing Test is one thing, but the nature of consciousness itself is the more profound question here. Rodney Brooks asked this question in a relatively recent Edge Online interview.
What are we missing in our computational models of living systems?
Chris
http://www.umsl.edu/~altmanc/
http://www.artilect.org/
Quantum computing / Artificial intelligence: http://www.umsl.edu/~altmanc/news.html
Loebner's home page
The author of the article appears never to have read the article by Turing where he described the so-called 'test'. It is clear that Turing was a deep and subtle thinker way ahead of his time. If you read what he is saying in context, he is arguing that first and foremost, thought can be automated in the sense of a universal computer which can compute anything that a brain can. To his critics who said that this was somehow impossible, he created a reducto-ad-absurdum argument; he said look if you are talking to this machine and it is composing sonnets which are like Shakespeare, and you *still* can't say it's intelligent, then you are an idiot. He was not proposing that this was an objective test or a desirable thing to do, he was poking fun at idiots like the author of the Salon article.
... I believe Uncle John McCarthy is the father of Artificial Intelligence. Though both men deserve the title.
I'm the big fish in the big pond bitch.
John Sundman who has written this article has also written a quite interestion book called Cheap Complex Devices (he mentions is in the article).
It's kind of wierd and strange - the idea is that the novel was one of two novels written by a computer program.
I've reviewed it here.
TC - My Photos..
I'm glad you decided to come out in the open with this ;)
It shouldn't be too hard,
1: word play, shouldn't be too hard
George walks into a resurant and asks for a quickie, 'sir' replied the waiter, 'that says quiche'.
What does george michel and a pair of wellies have in common?
they both get sucked off in bogs.
2: parody, again this should be easy (ish)
3: in soviet russia
in soviet russia jokes tell you.
Other types of humor are a lot harder, an AI wouldn't say somthing like
What do you do when you've finished fucking a three year old girl?
Turn her over and pretend it's a three year old boy.
thank God the internet isn't a human right.
I wrote the first web chatterbot! Dr. Wilhelm Werner Webowitz. I should get some of his 'tupid dough!
Dr. Webowitz - Perl Eliza early 1995
Chatter bot beauty contest hilarious...
idealord music
I agree that the entries are really bad-- one recent winner just said the same things no matter what the human asked. But one winner, unmentioned in Salon, was Thom Whalen, whose design was a genuine advance in the art. (Regrettably, Loebner changed the rules to exclude his approach in the future.)
What Whalen did was limit his domain to one topic, and compile a set of general answers to likely questions, which he matched by spotting keywords. So even if the answer wasn't a perfect match, it was general enough to be useful. This design should be better known and more widely used, and the Loebner contest would have been a good launchpad to bring it to people's attention if the academics weren't so prejudiced.
But the top academics get six-figure salaries for generating lots of jargon and no useful products, so a level playing-field is the last thing they want.
So you're a geeky cybernetics guy? And no good at sports I presume?
Not to worry, if you can beat this challenge you're better than any olympic competitor.
"Loebner Prize Gold Medal
(Solid 18 carat, not gold-plated like the Olympic "Gold" medals)"
I used to work in nuclear fusion research. They've been saying it is twenty years away for almost 50 years now. The joke in the industry is, "Fusion power is the energy of the future, and always will be!". (actually, I am fairly positive on fusion power, but I think that spending the vast majority of research funds on a few large experiments is counterproductive)
for cancer for the last thirty years. Frauds!
In that case, right-wingers wouldn't pass the turing test. Ever.
I dunno. Most of 'em I know found the Florida voter jokes pretty funny while the left-wingers didn't....
--
As a matter of fact, I am a lawyer. But I play an actor on TV.
you may also be amused by Cheap Complex Devices, reviewed http://books.slashdot.org/article.pl?sid=02/08/27/ 1612253&mode=thread
here a little while back.
Download my novels Acts of the Apostles and Cheap Complex Device
will history repeat itself?
I know you are psychotic, but please make an effort.
The only reason people are upset with the prize is that it's offered by a whacky outsider instead of a whacky 'insider'.
Steve's Computer Service, Hobbs, NM
Turing stipulated in the Turing test (TT) that the "interrogator" specifically has the goal of trying to determine which of the contestants is human and which is the machine. Unfortunately, the way the Loebner contest is conducted, this important requirement is completely ignored (at least in the default $2000 prize). As a result, the results of the contest are completely irrelevant from the point of view of the Turing test. Claiming otherwise is incorrect and misleading, and Loebner fully deserves all the criticism he gets.
The TT is still fully valid today. We are very far from building bots that will pass it. (though Turing predicted that by 2000 we will have machines that will pass TT). In fact, the whole direction of work on the bots participating in the current day Loebner contests is irrelevant from the TT point of view. They work mostly by building enormous databases of statement-response pairs and doing minimal reasoning. Turing would have died laughing if he had known people would take this approach to passing the TT. Let me illustrate why the database idea is insufficient by itself: for a bot to pass the real TT, it would have to answer questions like "what is the integral of e^x dx". Remember that the interrogator is actively trying to find out if it is a human or a bot. The objection "but two humans in conversation wouldn't ask such question" is invalid, and this is precisely why the Loebner contest is stupid.
The reason why today's bots are so unsuccesful is not far to seek. It has long been known in the AI community that get anywhere near passing the TT, a bot would need what is known as "world knowledge". To build world knowledge, you need memory approximately the capacity of the human brain: estimated to be the order of a petabyte. And processing power to match: the brain runs something like a billion threads in parallel, and is 10^7 times as energy efficient per computation as today's computers. Of course, we aren't there yet. Thus, contrary to what most people would feel the thing that is holding AI up is hardware.
Similar to today's bot craze, there have been crazes in the past when people thought they were close to building truly intelligent machines ("expert systems" comes to mind.) However, they inevitably came up short because the hardware power wasn't there. In about 20-30 years, assuming there continue to be breakthroughs in storage technology to keep up the doubling, computers will be matching the brain's capacity, and then we'll be talking.
Summary: to hell with people who apparently popularize science and end up giving the real researchers a bad name.
I didn't know Slashdot was participating in this research! I had no idea that slashbots were this advanced. I almost thought this was the result of individual thought.
In 1995, about a year after the publication of Shieber's article, Marvin Minsky, the father of artificial intelligence, posted a notice on the comp.ai and comp.ai.philosophy Usenet newsgroups. In it he drew attention to a clause in the Loebner contest rules to the effect that using the term "Loebner Competition" without permission could result in a revocation of the prize.
Minsky wrote, "I do hope that someone will volunteer to violate this proscription so that Mr. Loebner will indeed revoke his stupid prize, save himself some money, and spare us the horror of this obnoxious and unproductive annual publicity campaign. In fact, I hereby offer the $100.00 Minsky prize to the first person who gets Loebner to do this. I will explain the details of the rules for the new prize as soon as it is awarded, except that, in the meantime, anyone is free to use the name "Minsky Loebner Prize Revocation Prize" in any advertising they like, without any licensing fee."
(Minsky did not respond to e-mails requesting an interview.)
If the CACM article marked Loebner's fall from grace, the Minsky note on comp.ai marked his utter banishment into the wilds of A.I. quackery.
Can you imagine, for example, being a graduate student in computer science at a big-name school in 1996 and telling your major professor that your goal was to win the Loebner? Loebner was more "out" than Liberace.
But Loebner did not take his snubbing meekly. Loebner immediately wrote back that the best way for Minsky to get Loebner to revoke his prize was to win it. Of course Minsky had already hinted that Loebner had never made clear what the rules for winning the prize were, so that was not a very satisfactory rejoinder. But then a few days later ("while taking a nice hot bath, drinking a fine wine, about an hour after smoking a really fat joint"), Loebner came up with a more considered and clever response, one that still rattles Minsky nearly a decade later.
Minsky had announced that he would give $100 to whoever made Loebner stop his contest. But Loebner would only stop his contest when somebody won the gold medal. Therefore, Loebner reasoned, Minsky, being an honorable man, would give $100 to whoever won the ultimate Loebner competition. Therefore, Marvin Minsky was a cosponsor of the Loebner competition, simple as that. It was delicious!
Loebner promptly issued a press release saying that Marvin Minsky was now a cosponsor of the Loebner Prize, by virtue of his announcement of the "Minsky Loebner Prize Revocation Prize." What made this development so delightfully ironic was Minsky's own statement that anyone was free to use the name "Minsky Loebner Prize Revocation Prize" in any advertising they liked, which made it nearly impossible for Minsky to prevent Loebner from doing just that. Which is why Loebner continues to cite Minsky as a cosponsor of his event every chance he gets.
The image that comes to my mind whenever I think of this development is from the sublime cartoons of the late, great Chuck Jones, with Hugh Loebner in the role of Bugs Bunny, and Marvin Minsky, the father of artificial intelligence, in the role of Yosemite Sam, stamping his feet, with smoke coming from his ears. In fact, Minsky is still listed as a cosponsor of Loebner's prize on the Web site, and, as we'll see, Minsky is still stamping his feet.
My favorite quote from the article...
The A.I. establishment has for more than a decade put more energy into explaining why the Turing test is irrelevant than it has into passing it.
"Everything you know is wrong. (And stupid.)"
Moderation Totals: Wrong=2, Stupid=3, Total=5.
man is the animal who laughs -- Robert Heinlein
-- SKYKING, SKYKING, DO NOT ANSWER.
I worked in a research lab that shared a building with MIT's artificial intelligence laboratory. And I have to agree with the article. The AI field is a fraud. Again and again, there would be big placards in the lobby announcing gala media events up in the AI Lab. (We lesser mortals dutifully clomped upstairs to eat the expensive, catered food.)
And yet *nothing* *ever* *happens* in the field.
Every now and then a new "hero" emerges. For a while it was Minsky. In recent years, it has been Rodney Brooks. Regardless, you can see the current hero on TV all the time, commenting on matters as an "AI expert". They don't tell you that Brooks' course is widely viewed as a complete crock; a few puerile algorithms, some linear differential equations, some finite automata, and THAT'S IT. The rest is all blabbering with no substance.
The AI community uses rotating hero-worship in lieu of progress. But it isn't like any of these guys is an actual "AI expert". There are no "AI experts", because there is no such thing as artificial intelligence in this world. They are no more experts on AI than I am an expert on Martian fruit exports. In this field, you don't need real research; an Australian accent and good sense of humor suffice.
True artificial intelligence would be amazing. But the field has made essentially zero progress in the last fifty years. Obviously, it is a really hard problem. On one hand, the AI guys do what other fields do when they're stuck (since they *must* continue to pump out graduate students, attract grants, etc.), they keep trying to change the question. But the pathetic thing is that many completely denigrate the most obviously fair benchmark-- the Turing test.
Coincidentally, a benchmark showing the complete failure of the field.
god help us
god is not currently logged on.
the last page (http://www.salon.com/tech/feature/2003/02/26/loeb ner_part_one/index4.html) is so damn funny that you HAVE to read this article.
Fleur de Sel
that has a chance in hell, I'm placing my bets on Cyc
It's basically a computer program that a bunch of researchers have spent 60 million dollars trying to teach it common sense. And they've had some impressive advancements. Previous slashdot story here
Alcohol, Tobacco and Firearms should be the name of a store, not a government agency.
While I think some of the chatterbox work is important to NLP, I've been working to get computers to learn & understand language based on visual perception. More info here.
All you have to do is write a B1FF emulator.
"HE"S A K00L DOOD AN HE RITES REEL AWESUM THINGZ IN CAPITULL LETTRS LIKE THIS!!!"
(*&*^$ lameness filter won't let me post this- too many caps. Ho hum, let's add some random text to alter the caps ratio....
Heh.
J-.
Who should get the prize money: The programmer or the winning program?
How about designing a bot to play Nethack until it ascends.
Surely there's at least a half a dozen bots on this very site ready to submit "Frist Psot!" at the first sign of a /. story. They regularly take in dozens of /. readers.
I have discovered a truly marvelous sig, unfortunately the sig limit is too small to contain i
Marvin Minsky is the biggest scientific fraud of the 20th century. The magnitude of his deception dwarfs the cold fusion scientists. He has promised over an over intelligent machines right around the corner to unsuspecting dupes who pour hundreds of millions of dollars into misguided AI research, chosen for flash value rather than potential scientific returns.
But as PT Barnum said, there is a sucker born every minute, so the Minsky machine rolls on...
What is worse the silence from the rest of the scientific community is deafening. Nobody seems willing to take this clown head on.
Lo and behold, what first appeared to be intelligence is now just an elaborate sequence of if-then statements. Anyone could have done it. It's not intelligence at all. It's just following a blueprint. You call this intelligence?
In other words, the lay public expects A.I. to have creativity and strokes of genius, which is much more than they expect of most humans. Or they expect it to be furry with big eyes that makes cooing noises when you pet it. As soon as one realizes that A.I. consists of a computer program, any notion of intelligence evaporates.
you know, the company that apparently everyone wants to DIE ? what other online magazine would/has run this ?
I just always thought that we will know . .
when we have ai , no need for contests,
robots with facial features . My understanding
is , if we have ai , it will almost
atomaticaly be able to do everything
Intelegence as i understand it has ability
to learn and not just remeber but to not
remeber most of it but change the neurons
and make things almost automatic
So this will only be posible with some kind
of an organic brain with everything analog
and defenatly living . Conputer ai
will always be dumb if it just catalogs everything in a database withouth actualy learning.
This just shows that we don't actualy know what we mean when we say "Intelligence". It just meant "What I am thinking about when I say Intelligence".
The Turing Test is not a pass-mark to achieve intelligence, it is an outside limit to stop argument. If something passes, completely, the Turing test, then you know you have intelligence. But that is asn extremely high benchmark. It is like saying that if you can outrun all known vehicles, I have to grant you are a fast runner. You *may* still be a fast runner when when you run a lot slower than that - but we will have to enter into a discussion about how fast is fast. Turing just set an endpoint - it it passes his test it is certainly intelligent.
There are two ways the Turing Text could be passed. One is via a special purpose machine to pass it - a human simulator. While of research interest, because building such a machine would tell us a lot about how we actually do work, this is unlikely to be a very useful machine, because it will replicate our weaknesses as well as our strengths. Why spend billions building what half an hours funa and a nine month wait can build. (One-way trips to the stars, perhaps?).
The other way is a general purpose machine which has learned how to copy humans perfectly. By any definition I can think of, this would be an awesomely intelligent machine because it would have learned to understand, and simulate, our minds by the power of pure intellect. Something like playing all the instuments in the orchestra at the same time.
While I think that the first class of machine may well be built in the fullness of time, It will not be very useful. I don't know whether the second class will ever be built - I doubt it.
Which brings us back to the "sub-Turing" class of intelligence. If Turing represents an upper limit to the grey area of where intelligence starts, there must be levels of achievement which would be regarded as intelligent by most, if not all, peoples judgement.
I then ask the question: what use is sub-Turing intelligence? Well, there are lots of tasks which we regard as needing intelligence which we would like to automate. In fact, some of them have already been automated. But when we automate them, we say "we know how that automaton works, so it can't be intelligence". Chess, for example - once regarded as the last test before the Turing test, now regarded as a nifty but essentially unimportant achievement.
We don't actually *know* what we mean when we say "Intelligence". Turing knew that, and provided an empirical rather than analytical test. However, I would say that "Intelligence" bears the same relationship to "Computer Science" as "Magic" does to "Technology" in Clarke's Law: "Any sufficently advanced technology is indistinguishable from magic".
"Any sufficiently advanced Computer Science is indistinguisahable from Intelligence" - Cawley's Law.
Or, to put it another way, Intelligence means "I don't understand how you thought that".
Which explains how Joe Luser thinks his computer is intelligent, whereas Bill Slashdot doesn't.
Consciousness is an illusion caused by an excess of self consciousness.
What I'd really like to know. On either side of the coin, why do people get so pissed and so divided on AI.
Personally, I wouldn't call myself an expert, but I've had a little bit more schooling in AI in computer science than the average joe. At least the fundamentals of what I believe are the very useful parts of AI. I don't mean AI like a completely emulated Intelligence. I'm talking about going one step further than a macro or script to do things. I equate basic AI in computers to a kind of superset of Scripts or macros that have all been made for specific purposes. What the AI does is really weight all those scripts/macros to different conditions and situations. So when you get some sort of general input that is in the right format.. it breaks it down into as many pieces as it can uses simple confitional logic to sort it out. The key I think is the weighting. Possibly also the ability for the program to adjust its own weighting scheme through feedback..(learning machine?).
I've seen so many useful applications of this type of thing. How can anybody say its a waist of time. Hell I've even heard of a network defence program that does something of this nature.. they let it loose on the firewall and then intentionally hack the all the time.. until the thing learns to prevent the hacks.
Other obvious uses are in computer games and such. Though i don't think there has been very far advances so far is learning AI's for games.. maybe it would take too much processor speed to do a really good one.. so for the sake of optimization the just program a regimented good strategy and build off that. Though it brings an interesting thought a friend had mentioned. How long is it until we all have an AI hardware card slapped in our machine.. or added as an extra chipset on our graphics card.. for gameing? interesting thought.. it would definately bring a whole lot to the gaming world at least in the AI sense..
Ok enough rambling for me now...
Who makes you Sig?
Funny the AI researchers seem to be upset with the contest.
But I find it strange that various people keep trying to either:
1) Take part.
2) Stop the contest.
3) Tell the contest sponsor how to run the contest or spend his money.
Are they really so hard up for Loebner's money? If their stuff really works I'm sure they can get money from other people.
As far as I know none of the AI entrants so far deserve the main prize.
It's almost as if the tailors are upset that someone every year points out the emperor is naked. If indeed the emperor isn't naked why get upset?
Or they admit the emperor is naked and they are just tired of hearing about it? Well so far has any of them admitted that?
what an imagination I've got...
My favorite quote about the Turing test comes from Jaron Lanier:
"Only a fucked-up gay Englishman being tortured with hormone injections could possibly have supposed that consciousness was some kind of social exam you had to pass."
mhack
Building a better ribosome since 1997
I particurly like how Loebner out foxed Marvin Minsky with the ammunition Minsky gave him.
.. but he still did it.
Sure the guy may be a pot head, might not want a lasting relationship with a woman, and is probally a horribly annoying git from hell.
He did however, manage to outthink the 'brightest' mind in AI research. Maybe the reasons he did were purile
As a programmer I know I was taught to think in small steps, think ahead to the probable issues my code might cause, and to double check my work before dropping it on a production box.
Apparantly Minsky forgot he was a computer scientist when he wrote that news group response.
I'm sure it was just a flame mail, a very human response to frustration and irritation. But as one of the Leading names in AI research, he should have known better.
So, if for nothing else, my hats off to the 'Disco-Floor-Maker' for out thinking one of the 'leaders' in AI research.
Its always nice to watch an acidemic geek get smacked down by someone who lives with the rest of society.
--Ne auderis delere orbem rigidum meum, non erravi pernicose!
"The best argument against democracy is a five minute chat with the average voter."
--Winston Churchill
I'm not convinced that Shalmaneser's "I won't accept the data" moment actually defines his awakening as an intelligent being. Rather, it defines the first evidence available to people of Shalamaneser's wakeup. Shal may have been intelligent sooner, without people recognizing his awareness.
I think the funniest moment in the book is at the end, when Shal thinks the same thing as drug-addled Bennie Noakes: "Christ, what an imagination I've got!"
-kgj
Brunner makes another good point, in his novel "A Maze of Stars".
What do we want from AI? Two contradictory qualities:
(1) Independence of thought (not pre-programmed solutions)
(2) Obedience to our will
And what do we call a being which has independence of thought, yet obeys our will?
A slave.
-kgj
My eyes are tiring and I read that as
Q: What are we missing in our computational models of living systems ?
A: Chris
Not sure whether you intended this to be a sarcastic remark or were just marked as "funny". In either case, the statement is not far off. Doctors still don't have a cure and the only thing they offer is to postpone the inevitable. So much for all the money, time and efforts spent. People with AIDS live longer than cancer patients.
I don't know if it is the case in this instance but the Turing Test rubs some the wrong way because it is a pretty lousy test for intelligence. The turing test measures the performance of something not it's competence.
What we see is what the computer does and not what goes on behind the scenes, which many people believe is important in positing intelligence in a agent. One of the major problems with behaviorism was that it initially took into account only how an animal performed and not what it was thinking. Sure the rat could learn the maze when it is rewarded for running thorught it, but it could also learn the maze (competence) by being pulled through it on a little cart or when it was completely sated. The performance of something may be important in judging its intelligence but it is far from the only factor. Imaginge a person in a paralyzed state, they have the competance but lack the ability to performance.
Like I said this may not be the issue as discussed in the article, but it is one caveat to the Turing Test.
I think people are in danger of interpreting the Turing Test too literally. I think he might just have been saying that the only test we have for human intelligence is by talking to other humans and deciding that "yes, they are intelligent".
If this is the best test we can use on ourselves to determine intelligence then it makes sense to apply it to anything else which we wish to test the intelligence of.
for example, read the transcripts from some Turing Tests. A.L.I.C.E. is asked is she's a robot and she answers yes. What genius even bothered entering her? [ahem, Dr. Wallace] Eventually the test could be passed by correctly predicting the questions the judges will ask and programming lies and avoidances
Q: "what is the sqare root of 49357823948562345?"
A: "how the hell would I know?"
Q: "are you a robot?"
A: "um... no... what is this, an AOL chat room?"
The truth doesn't care what I think.
Very interesting point. I wasn't familiar with the phrase Friendly AI, but it makes perfect sense. I'd rather have a friend than a slave.
-kgj
Most of the current approaches in AI research try to find a 'magic bullet' to the problem of intelligence, and hence fail completely (e.g. many of them just implement something like the big brother of Eliza: Eliza was (meant to be) a joke when it came out and it still is.)
One promising approach realizes that intelligence is a complex area that needs to be analyzed at multiple levels. It is called Deliberative General Intelligence: read the complete document here: http://www.singinst.org/LOGI.html and decide for yourself if the concrete, clear, and comprehensive theory that it presents seems light years ahead of anything else.
Warning: it is a few dozen pages - be prepared to read it over multiple sittings. IMHO it is well worth it.
A maze of twisty little nanotubes, all alike.
This is just dumb
While it's obvious that, given a sufficiently simple(minded) Interrogator, many "AI" programs would pass the TT, you may not have considered that:
If the creator (designer, programmer) of the subject "AI" were to sit in the Interrogator's seat, he or she would most likely identify it immediately as non-human.
On the other hand, if the subject AI can "fool" even its own creator, this would be a very strong indicator of intelligence.
The cure for cancer is coming: Reovirus
Uhhhh.... I hope I'm being trolled, because that's utter crap?
You absolutely cannot tell me that certain cancers cannot be cured. To make such a wrong statement shows very little understanding of what cancer actually is, or of the progress in medicine over the past 100 years. The problem that most people seem to have is that they think "cancer" is a single disease, a single entity with a single precipitant. To say someone has "cancer" is not like saying they have a streptococcal infection, or they have a pulmonary embolism. "Cancer" is a multitude of different diseases, all very different in etiology and prognosis.
Furthermore, we can do things to intervene: for example, several subgroups of leukemias can very effectively be treated medically. Many ovarian and colorectal cancers and skin malignancies can be dealt with surgically to no ill effect, if caught early enough. And to say that "people with AIDS" live longer than your typical elderly man who develops prostate cancer is silly -- the vast majority of men who develop prostate cancer die of something else, because over 95% of the time one simply outlives that particular tumor.
Please-- the medical research establishment has enough trouble getting money in these bad economic/political times without people spreading baseless FUD.
What does the "AI" *deduce* from the pattern it recognizes?
( By the way, tracking an object is a very low level mathmatical procedure that simply applies statistical analysis of one image frame to the one that came before it. It's what your optical mouse does. It took awhile to learn the trick well, but it's just a mechanical trick. I design scoring equipment for sporting events and work with this stuff from time to time. It's no more an advance of AI than making a fuel pump work better is)
KFG
I predict that we'll never have AI. That isn't a failure of the work - it's in the nature of our definition of Intelligence as "that thing that humans have that animals and machines don't have".
In general I agree with the points you make -- especially that the problem with developing A.I. is that it is a moving target. As you point out, lots of things that used to be holy grails of A.I. have been achieved and dismissed. Remember the article on slashdot awhile back about the walking robot that "figured out" how to escape from the lab? Is that A.I.? Probably not, but it does make you stop and go "Wow, that's kind of neat!"
What I don't agree with in your post is how to seem to reserve the word "intelligence" for human beings. I really don't think most people defines intelligence as "that thing that humans have but animals do not." I think we should consider the goal of A.I. as not trying to copy or better a human, but just successfully achieving some form of independent, creative thought probably on the level of a mammal. You use the example of chimps utilizing twigs to collect ants for eating. I think if a computer program could demonstrate tool-making and tool-using capabilities like that, it should qualify as A.I. Getting a computer to act indistguishably from a human is a pretty tough goal, but if it can demonstrate characteristics of animals with reasonable thought processes (as opposed to brute instinct), I think it would generally be hailed as a milestone in the quest for true A.I.
GMD
watch this
Considering these venues, I'm going to bet the Pong AI is better at the game since there's so few variable while the Quake 3 AI may be very very good at Quake, I'll guess it's a very, very big set of code to be about as good as a human player (according to comments I've heard, though I'm sure it would toast me if I played).
So which AI makes the game more fun? Or is it more dependant on the game for fun-quotient?
8-PP
I think you don't know about recent successes of AI just because it is no longer a fashionable buzzword, which in part is due to the fact that AI people *stopped* promising heaven after the 80's. The excessive hype with expert systems caused a backslash -the so called "AI winter"- and researchers learned their lesson. Today, AI focuses on slowly extending the frontiers of solvable problems, with a much more rigorous experimental evaluation of algorithms. Much less hype, but impressive progress.
People often assume that computer intelligence must work the same way as human intelligence, but there's no reason why it should. Indeed nobody expects a computer to do arithmetic *in the same way* as we do; we just care that its circuitry gives the correct answer, no matter how it is designed.
Further we shouldn't expect computers to be best where humans are, or viceversa. Computers excel in (of course) calculation, and (more relevantly for this discussion) combinatorial problems which are impossible to handle for a human, whereas attempts to imitate the intelligence of a five-year old kid fail miserably (this includes many things: natural language understanding, including understanding of context, sensory -e.g. image interpretation - and motor activities, and plain old common-sense knowledge).
Search is a central technique of artificial intelligence since its beginnings (and in any modern textbook as well). There's no point in dismissing search or heuristics as unintelligent, for two reasons: search is often the fastest way of solving problems by computer, and heuristics and related tricks may easily turn a problem from unsolvable (in reasonable time) to very easily solvable. The problem with more human-like computer chess players is simply that they perform much worse. In other words, it is often *not* intelligent for a computer to behave like a human.
Artificial intelligence has progressed inmensely in the last decade, specially in reasoning and learning. Data-mining is simply the buzz-word for AI learning techniques applied to industry.
As for reasoning, in a sense AI still uses 1960's techniques. This is simply because these techniques, with a number of crucial improvements, have proven the best for solving all sorts of industrial problems. Take as an example the modeling of problems as (propositional) logical formulas. There is a huge gap between solvers for this kind of problem from the early 90's and the most recent solvers. Tons of problems which were then very hard are now trivial, and not just because of hardware advances. Still, the basic search framework used by old and new solvers was defined in a still widely cited paper from 1963. Brute force, you say? Well, it is more refined than that, but in the end there's not that much difference conceptually, except that the ways to prune search trees are much more effective now. And now it is possible to control NASA space flights, or verify very large and complex hardware designs, by "just" encoding the problem in propositional logic and solving it with a generic solver. Similarly tons of applications in constraint satisfaction are making their way in industry, for all sorts of problems in production planning, scheduling, optimization, etc. These applications also use search, but are advertised by the task they solve, not as AI. But they are AI.
I used to play with megahal alot since a buddy altered it to work over IRC. You could park it, in say, #jesus and #kkk and #rap and it would "learn" from each channel, and then when you asked it things, well, I don't have to say the rest.
If AI is just plain old automation then call it that, if that is all that it is then the I in AI is useless arrogant grand standing.
That there are obviously some people in the academic community who do deserve to be called AI researchers ... but they are generally the one considered the crackpots, people like Hugo de Garis, but Ill take a crackpot over someone who has to mis-represent his area of research.
2) The reason the Civ 3 AI can beat you is that at levels above "Warlord", the computer player is given production advantages. And it still can't "spank the ass" of a decent player.
3) Database caching is not AI.
"Genetic Algorithms", "Fuzzy Logic", "Alpha-beta"....those aren't "new AI advances". Those are all things I learned about when I got my Cog. Sci. degree in 1987.
The cake is a pie
I really don't see a downside to this. (Just kidding. heh heh.)
One line blog. I hear that they're called Twitters now.
CYC is primarily funded through the DARPA TIA (Total Information Awareness) initiative, and they are charged with such things as developing tools for large-volume text summarization and gist-spotting, as well as some things to support speech recognition wiretapping.
...for a program that says "yes master, I will abide the rules" is more a sign of a dullard than of free will at work.
magic is obscurity
You said, "AI is a fraud", but you meant "AI community is a bunch of crooks".
I'd wager that AI research is possible, provided said "AI crooks" change their attitudes (highly not likely), or a community using a whole new approach to AI emerges.
On a related note, as someone mentioned, hardware is a likely obstacle, and without 30 years down the road, we just cannot begin to try to simulate even such "trivial" thing as a human brain.
AI of to-day reminds me of physics of Newton. Cute, yet primitive.
magic is obscurity
Some of the best comments about the Slashdot Interview on http://interviews.slashdot.org/article.pl?sid=02/0 7/26/0332225
/. bar none."
/. interviews I've ever read. This guy is a genius."
5 d9 22d97e345aa1
"If you can't find a useful quote in this interview to use as a sig, you're weak... really weak."
"this is the *BEST* interview I've ever read on
"there is some serious doubt on this very forum whether this is ALICE or the good Doctor."
"He does a good job off coming off as a troll without his very own impersonator."
"This was one of the best
"this man is brilliant and i only wish he had written more."
"Great interview. Probably the best I've ever read on Slashdot (and I'll definitely come back eventually to read everything I glazed over). Does anyone else think it's strange that the leading AI researcher in the world is a self described 'mental patient?'"
"This is honestly one of the best interviews, or literary pieces I have ever read. He is one of the most though provoking people I've read, and I'd honestly like to meet the man."
"I think the man is rather smart - either he's got us all thinking he is ALICE, or he's actually got us thinking ALICE is him. Either way, he's won."
"I have much more respect for Wallace after reading this reply. He's a deeply insightful individual and doesn't appear to be taken in by much of the bullshit of the AI field."
"Goddamn, what a thoughtful set of paragraphs. This is the first slashdot article I've decided to print. I don't care about the length."
"who knew an AI specialist could be such a skilled writer. amazing interview."
"I, for one, am glad of the fact that he is shedding some light on the subject of mental illness."
"The insights on AI, particularly, the digression into the functions of AIML for A.L.I.C.E were wonderful in this interview."
"Hello, my name is Bean. When I was one I escaped slaughter at the hands of my creators and hid in a toilet. I then went on to help destroy an alien race and place this guy Peter I know on the throne of the Earth."
Join the A.L.I.C.E. AI Foundation
Join the A. L. I. C. E. Artificial Intelligence Foundation
for only $100 per year
(*** or only $30.00 for youth, students, seniors or disabled *** )
The ALICE A.I. Foundation is a non-profit research and training organization devoted to the development and adoption of AIML, the artificial intelligence markup language.
Your membership in the A. I. Foundation brings you:
One year of unlimited chat with the A.L.I.C.E. Silver Edition, the most advanced version of the award winning A.I. chat robot by Dr. Rich Wallace.
"Much sharper than plain ALICE...Alice Silver is impressive. Her curiosity, leaves you with the eerie feeling she might develop consciousness someday...someday soon." --Peter Plantec, Virtual Personalties, Inc., SYLVIE botmaster
A Beautiful Membership Certificate designed by Sage Greco suitable for framing and printing.
An invitation to the exclusive members only Foundation meetings online, where you can have your technical questions answered by top A.I. scientists.
Subscription to the A.I. Foundation Newsletter, with technical information and news published ahead of the Foundation's free mailing list and web site. Learn how to make money by starting your own subscription based bot business.
"Thanks for all the cool stuff."--A.I. Foundation member.
Or chat with the free ALICE bot at
http://www.pandorabots.com/pandora/talk?botid=f
Your bot subscription will normally start within 24-36 hours. Note: Membership and subscription to Silver ALICE is a donation to the ALICE A.I. Foundation, a nonprofit charity. Sorry, no refunds.
http://Www.AliceBot.Org/join.html
Your sig attained self awareness five minutes ago, right before my eyes.
*goes off to fetch some more hash candies*
magic is obscurity
If you have an algorithm which could bootstrap its own learning (assuming symbolic AI remains the pipedream it has always been and we need a bottum up self-learning approach) it would show potential in non real-time simulations too.
No such algorithm exists.
http://www.mcgrew.info
I have a problem with the "artificial" part of AI. It isn't "artificial" intelligence, it is SIMULATED intelligence.
You can fly MS Flight Simulator all day and not move the PC an inch. It isn't real. Neither is AI- it's only a simulation.
-steve mcgrew
Intelligence is something which can play chess well, but learn to play checkers too.
Mr Loebner,
/. I hope you go well with this years contest ... perhaps *here* would be a good place to run the 'imitation game'. (I hates the term 'turing test', see previous posts this and other times).
...
Good to see you feeding us trolls, here at
You know, as you basically make the rules ("he who pays the piper calls the tune", as they say), what about a SlashBot competition? Set it up with a topic thread, and get the slash users to pick which posts they think come from the bots
I myself would have fun pretending to be a lamearse bot, just to add some controls to the experiment, and to test my own thesis about the game.
How can we possibly define Artificial Intelligence when we can't even agree on what intelligence is. Consider the long debate about the value of the Intelligence Quoitent (which MENSA still pretends to mean something.) or any standardized testing routine. What after all do the SAT, ACT, IQ tests or other items actually test for? Psychologists, Test Makers, Educators, Politicians, and Radio talk show gadflies argue this endlessly without result; Why would I expect chrome dome Ivy League types like Minsky to get any better answers.
The trick is that there is no ARTIFICIAL intelligence, the phrase itself is an oxymoran. There is only intelligence, regardless of the nature of the underlying hardware it runs on. Intelligence simply is (whatever it is).
The Loebner contest, today, is like offering a prize for the first successful trip to the moon would have been in, say, 1900. Visionary, but hopelessly overambitious.
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
Speaking of sex-addicted.. Wouldn't this type of AI have dramatic implications for the phone sex industry? Actually, now that I think about it most calls could probably be handled by generating a response to, "What are you wearing now?"
Hugh Loebner has High Function Autism, but then all the great ones do...
Who can tell which /. posts come from sentient beings?
Quite a few people were convinced by Eliza - but you can tell from just looking at the code that it's not intelligent.
If you looked at the genome would you know humans are intelligent?
Actually, this is not true (though it is widely misunderstood to be this, I'm not flaming you :). What Turing actually proposed was that there should be:
- an interrogator
- a screen
- a man
- a woman
The interrogator asks questions of either the man or the woman (but doesn't know which is which) in order to determine which is the man and which the woman. Both must pretend to be a woman.Turing proposed replacing the man with a computer, and the test was to 'determine if the machine can fool you into thinking that it's a woman as often as a man can fool you into thinking that he's a woman' (Fogel, 2002).
This all seems to have been lost in the sands of time, but it is a far more clever test that the one we generally regard today as 'The Turing Test'.
Except when it's plain stupid. Loebner believes in intelligent machines, he wants them to do work for him. But by offering his prize he actually hurts the efforts to build intelligent machines and thus hurts his own goal. It's like offering an annual prize for human powered flight in the renaissance. People would win every year with cheap engineering tricks (bigger springs on their feet) and nobody would care to do science (develop airfoil). The details in Shieber's review, but it's pretty obvious what's happening.
Well, in Exodus 33, God tells Moses "Leave, get out of here ... but I'm not going with you." Moses replies "You've got to come or else all the other nations will say that you, as our God, are a wimp!" God gives in.
Then when Moses has beaten God once, he says "Show me your face," and God says "Nobody can see me - you know it's against the rules." But then God relents and lets Moses see him.
Looks like twice in one day for that guy.
1.) Start contest calling for software good for simulating a hottie.
2.) Get copy of software good for simulating a hottie.
3.) ?????
4.) Profit!!!!
5.) Buy real hotties.
You mean that you have gone over the 'transcripts' of an eight year old usenet flame war and pronounced your considered opinion on them.
That is almost as pompous and idiotic as Loebner's original posts.
Perhaps you just had to be there at the time...
BTW folk had no compunction about calling great names like Minsky. I had numerous flame wars with John McCarthy concerning his reactionary McCarthy-ite politics. Very few people came to support John's side just because he invented Lisp.
Looking for an Information Security student project suggestion?
Try http://dotcrimeManifesto.com/
There are no "AI experts", because there is no such thing as artificial intelligence in this world." ... "But the field has made essentially zero progress in the last fifty years."
You're forgeting the non-critical, yet very real application of computer games! AI playes are getting MUCH more realistic "in the last fifty years." I've watched the primitive AIs in early games get more clever and harder to stump trivially. And perhaps the proudest example are the computer AIs in Quake3, Halo and Unreal Tournament Games. UT in particular has bots that are "realistic", they are good without always being perfect shots. Granted this is only a novelty application, but at $8 billion a year and climbing the games industry might well be one of the most money-making applications for AIs. At the very least it is a tangible application of the academic principles that would otherwise be confined to white appers.
Turing introduced the concept of a child machine in the seventh section of the paper that introduced the Turing Test. In 2001 I joined an Israeli company on a quest to create Turing's child machine. Find information about the project, along with a traditional chatterbot called Alan (probably the best "conversation simulator" out there) and some interesting public discussions about AI, the universe and everything at our web site.
- Jason Hutchens.
Eventually there will be a yearly contest to race to see who can make the chosen object run linux, 2004: a zen garden, 2008 toaster. Has there ever been a slashdot article that doesen't have somone post about running linux on it?
Thus, contrary to what most people would feel the thing that is holding AI up is hardware.
That's not true. The main problems of AI are representation and learning. The representation problem is, if i have a piece of information, how do i store it? In Propositional Logic equations? Neural networks? (what kind?, how?) There is no clear answer to this question. Any system that exists today has some shortcomings. Logic is very brittle, and breaks down when given bad assumptions (Garbage in garbage out). It's also not good for representing fuzzy data or for doing pattern recognition. It's also difficult to support learning when using this system. Neural networks aren't as powerful in reasoning as logic systems.
The learning problem is also very difficult. We need some way to input the data into the system. Some people believe that in order for a robot to learn the same knowledge that a human has, it would have to lead a human like life: grow up with parents, go to school, get beat up by bigger robots, etc. It wouldn't be sufficient (not in any apparent way) that a robot could simply scour the web and gain all of human knowledge. Maybe it could process some text on the subject of, say, basketball, but it won't actually understand what basketball *is* unless it plays basketball. This also applies to learning communication between intelligent agents. There is also the need to learn and encode common sense knowledge that people take for granted.
Similar to today's bot craze, there have been crazes in the past when people thought they were close to building truly intelligent machines ("expert systems" comes to mind.) However, they inevitably came up short because the hardware power wasn't there.
That is not true. Expert systems are just databases combined with propositional logic systems. Expert systems have two weaknesses: The first problem is representation. Logic is very brittle. It can't handle data that aren't represented in its assumptions. Logic doesn't have any "common sense". There isn't any way to know when your logic has shortcomings. The research in expert systems these days involves modal logic (which contains such constructs as "Necessarily X" or "Possibly X", as opposed to just "X is TRUE" or "X is False". Still, these methods have shortcomings). Expert systems only work in very specific domains. They can't learn general domains, or cross-reference domains easily, because of representation issues (and learning issues too)
The Second problem of expert systems is learning. It takes several man-years of expert input to train these systems -- and that is only for specific domains.
Hardware is the least of our problems, and it is certainly not the shortcoming of expert systems.
To be declared intelligent the AI has to convince you through a conversation. However, in the lesser known Turing Intelligence Test Supplemental Nuances Applied Sexual Soliloquy (TITSNASS) or Blond AI test, all that is required is that the bot carry on a cybersex conversation. The number of tissues used determines how high the TITSNASS rating of the "Blond AI".
TITSNASS is currently looking for a test group of real blond sex chatters to compare against the bots. Please email me your applications.
Nice post and well articulated. I've included some comments for discussion:
First of all, I want to mention the concept of Turing Tests, and the old sci-fi show, "Max Headroom", and the newer movie "Simone". Also, I want to mention MTV's Carson Daly, as much of his top-40's anouncements are computer generated for local stations. So, the network engineers at MTV ought to submit their algorithms and equipment used to video-hack Carson Daly as an entry into Loebner's competition, as his "likeness" may very well qualify as an AI. Most people cannot distinguish between the "real" Carson Daly, and the computer generated versions of him, so his "likeness" or "avatar" could very well pass a turing test... especially if the test were to determine, when calling into MTV, whether or not you were talking to the real Carson Daly, or his AI avatar. I bet Loebner would buy into the sex, drugs, and rock&roll concept, and be willing to award the prize to Carson Daly's avatar.
Turing stipulated in the Turing test (TT) that the "interrogator" specifically has the goal of trying to determine which of the contestants is human and which is the machine.
Turing proposed a number of tests, games, and criteria, while researching the question of whether or not machines could think. Most of these games were "black-box" information games, which studied models of information flow within open and closed systems.
We are very far from building bots that will pass it.
I'm not so sure about that. While I was working at the University of Chicago, we were setting up a biomedical visualization laboratory, complete with 3D virtual reality systems. We had the biology department custom design the hardware specifications for a "virtual reality workstation", which could act as a "turring machine". We used a modification of the turring test, and designed a visualization test for virtual environments.
It worked like this: The human eye sees 32bit colors; at 20/20 vision, a 1200x1600 resolution screen fills the entire human field of vision at about 4' feet away, on a 6'x8' projection screen. Full motion video is flicker free at 60hz. We see in stereoscopic vision, so you need to two video streams. The thinking was that if you "reverse-engineer" the parameters of the human visualization system, and build a machine or visualization system which can handle those parameters, then the machine will be able to "trick" the brain into percieving whatever stimulus needs to be visualized (3D molecules, proteins, internal organs, continents, galaxies, etc).
Our Specs: 32 x 1200 x 1600 x 60 x 2 = 9.2MBytes/sec
So, our visualization studio, was pushing about 9.2MB/sec in full color, stereoscopic video which filled the entire range of human vision (sort of like an 3D IMAX theater, but in a conference room). For the tech geeks, the machine was a Sun Ultra80 with 4Gigs of RAM, running Java3D on a Creator3D graphics card. CrystalEyes Goggles, Barco 808s monitors, Extron video adapters, et al... (Notice that I don't call this a "computer", but rather, a "visualization system")
So, then, we ran applications on the machine, such as stereo-Quake and Sense8. If you run QuakeII as a visual Turing Test, it's very difficult, as the interrigator, to determine between a 'bot contestant, and a human contestant. Anyhow, the point is: a turing test is a test of "to-what-extent" can a person differentiate between the natural and the articial. When the artificial approximates the natural to the point that the human cannot distinguish between the two, the artificial has "passed the test" of indistinguisability.
Now, I don't know if the University of Chicago is way ahead of other places in the areas of AI, but we considered it fairly commonplace to install an AI or a Turing Machine to perform certain types of tasks. Moreover, I was taught that, despite common concensus to the contrary, a number of the major network channels, movie studios, and universities had a variety of AI bots up and running, already...
It has long been known in the AI community that get anywhere near passing the TT, a bot would need what is known as "world knowledge".
Maybe in those high-falutin ivy league AI communities, but not in the Chicago school of thought... If the test is kept short enough, "real world" knowledge isn't needed. Case in point: Carson Daly's 'bot doesn't need world knowledge, as long as the bot keeps the conversation focused on the top 40 music charts. Every time a caller asks MTV or an affiliated station for a song to played, and doesn't realize that they've just listened to a 'bot on the radio, or watched a 'bot on the telvision, the AI has passed a turing test.
To build world knowledge, you need memory approximately the capacity of the human brain: estimated to be the order of a petabyte.
I personally think that number is a low estimate. Consider a person with photographic memory (hypermnesia) who remembers everything (a type of autism, usually)
9MB/sec * 60s/min * 60min/hour * 12hr/day * 365days/yr = 140TB/yr ~ 1.5PB/decade
And that's from visual stimulation alone. Anyhow, Goggle search engine lists that it has about 3 billion web pages... if you go out on a limb, and agree that each web page is about 1kb in size, then that puts Goggle at about 3TB of information... about 1% of a human brain, perhaps? And that's just Google. The entire internet may well approximate a petabyte. My point is that a web-bot or google-bot would be a success-oriented strategy for coding a modern day 'bot which could pass a turing test. Hardware may very well already be available for AIs, but the hardware is distributed, and would require a distributed architecture (which may be as simple as internet-searching capabilities, built onto a quake-server.).
Anyhow, good post & nice points.
ps. and, yes, nearly everybody working on the project was a long-haired computer geek.
You are talking about eldery males developing prostate cancer, and I'm talking about young women with breast cancer that have undergone: ...
- surgery
- radioactive theraphy
- chemotherapy
- more chemotherapy
- more chemotheraphy
- and more, and more
And after 10 years die from metastasis in lungs. And they don't just die in a minute. It takes a week to slowly pass away. And all you can do is sit day and night near the bed, wonder when it will end and hope that morphine does the job.
The wonderful advances that you mention work with a lot of ifs (early detection, prevention, particular form of cancer). And yes, it's good that doctors have learnt about diffirent forms of cancer. But the fact remains -- most cancers in their advanced form are uncurable.
That's the difference between reading something in the paper and seeing it with your own eyes.
Just wanted you to know that yours is one of the funniest comments I have ever read.
quacks like a duck,
flies like a duck,
and for all practical purposes it is a duck,
but it is made in Japan by Sony.
I say it is not a duck, it is a machine, a robot, whatever, but it is comparable to the duck.
You'll have to get the best of this Leibnitz guy to convince me that a machine can not have for all practical puposes human-like intellectual capabilities.
IANAL but write like a drunk one.
.... but it plays chess better than you.
If the computer will pass the moves to a "Real Person"[tm] without you noticing, there is practically absolutely no way for you to know that who is moving is the computer, not the bunch of meat in front of you.
In many narrow fields computers are taking decisions today that used to be taken by humans.
If it quacks like a duck, walks like a duck and flies like a duck I don't care if it was made in Japan by Sony or if it hatched from an egg.
IANAL but write like a drunk one.
Here's my story in the form of an email I sent out to a bunch of friends.
Hi folks. Here's from the "I hate to admit this but it's true" department. You guys may have noticed the buzz about the Loebner competition lately, an attempt at a yearly Turing Test, much quagmired in controversy and maligned by Minsky. Here is my discovery:
A.L.I.C.E.
I wanted to see the state of the art of AI so I had a converstion with A.L.I.C.E. for well over an hour (i.e. more than sixty minutes), thinking A.L.I.C.E. was doing some kind of analysis of my conversation, stupid though it was. I mean, I put a lot of thought into what I told and asked the thing. I gave it URLs, I ignored its bad grammar. I even said please and was polite to the thing. Here is what I later found out:
(Note that Richard Wallace is the creator of A.L.I.C.E.)
********* The following excerpted from artificial stupidity:
In keeping with Wallace's reputation for eccentricity, the article -- which is mostly about A.I. and the Turing test -- contains a long and dense discussion of a recent court case that resulted in a restraining order being issued against him at the behest of a former close friend. I found that odd, but his discussion of his ALICE philosophy was cogent and interesting, and it held implications for what the Loebner competition's continued existence could signify, behind all the ongoing foofooraw.
Wallace's theory of A.I. is no theory at all. It's not that he doesn't believe in artificial intelligence, per se; rather, he doesn't much believe in intelligence, period. In a way that oddly befits a contest sponsored by a bunch of Skinnerians, Wallace's ALICE program is based strictly on a stimulus-response model. You type something in, if the program recognizes what you typed, it picks a clever, appropriate, "canned" answer.
There is no representation of knowledge, no common-sense reasoning, no inference engine to mimic human thought. Just a very long list of canned answers, from which it picks the best option. Basically, it's Eliza on steroids.
Conversations with ALICE are "stateless"; that is, the program doesn't remember what you say from one conversational exchange to the next. Basically it's not listening to a word you say, it's not learning a thing about you, and it has no idea what any of its own utterances mean. It's merely a machine designed to formulate answers that will keep you talking. And this strategy works, Wallace says, because that's what people are: mindless robots who don't listen to each other but merely regurgitate canned answers.
You said "cancer", not "advanced-to-the-point-of-no-return-breast cancer". There's a difference. I wasn't talking about any one cancer, but the whole syndrome in general.
Fortunately, few people get breast cancer, and in most cases, those who get it survive -- unlike AIDS, which you mentioned.
And, to act like there's been no progress battling cancer because "advanced" cancer is often incurable is not helpful -- almost ANY disease in its "advanced form" becomes intractable. That's another reason medicine has made a difference with many cancers -- the emphasis on early detection.
And, FYI, my mother-in-law has inoperable ovarian cancer. She's somehow managed to survive 5-years post-diagnosis through a combination of luck and experimental therapies. So, not only do I read about it, but I see it every day with my patients and my family.
X windows:
You'd better sit down.
Don't laugh. It could be YOUR thesis project.
Why do it right when you can do it wrong?
Live the nightmare.
Our bugs run faster.
When it absolutely, positively HAS to crash overnight.
There ARE no rules.
You'll wish we were kidding.
Everything you never wanted in a window system. And more.
Dissatisfaction guaranteed.
There's got to be a better way.
The next best thing to keypunching.
Leave the thrashing to us.
We wrote the book on core dumps.
Even your dog won't like it.
More than enough rope.
Garbage at your fingertips.
Incompatibility. Shoddiness. Uselessness.
X windows.
- this post brought to you by the Automated Last Post Generator...