Slashdot Mirror
New Contestants On the Turing Test
vitamine73 writes "At 9 a.m. next Sunday, six computer programs — 'artificial conversational entities' — will answer questions posed by human volunteers at the University of Reading in a bid to become the first recognized 'thinking' machine. If any program succeeds, it is likely to be hailed as the most significant breakthrough in artificial intelligence since the IBM supercomputer Deep Blue beat world chess champion Garry Kasparov in 1997. It could also raise profound questions about whether a computer has the potential to be 'conscious' — and if humans should have the 'right' to switch it off."
and see if it complains, first. If it does, then call me back.
Are they really *thinking* or have the programmers just done some tricks to make it seem that way.
"Teaching to the test", so to speak.
If we don't have the right to switch off a conscious machine (one that passes the Turing test) does that imply we have the right to switch off a human who fails a Turing test?
The purpose of (strong) artificial intelligence isn't to trick humans somehow, it is to figure out how our mind works. What is the algorithm that powers the human brain? No one knows.
Who cares if contestants can be tricked by a computer? Who cares if some computer can calculate chess moves faster than any human? None of this helps us get closer to the real purpose of AI, which is why they call it weak AI.
Qxe4
Why would it raise these questions? I don't think anyone would disagree that computers are far better at matrix algebra than humans could ever be, why isn't that the test? The ability to invert matrices differentiates from the other orders more so than language does anyway. Why this arbitrary test? It doesn't seem to have anything more to do with 'consciousness' than an ATM does. I'm not trying to discredit the hard work and progress here, but jumping to consciousness is probably not going to happen in software.
"... and if humans should have the 'right' to switch it off."
GROAN.
Whale
It could also raise profound questions about whether a computer has the potential to be "conscious" -- and if humans should have the 'right' to switch it off."
Maybe in the esteemed opinion of vitamine73 it will, but if you knew anything about how artificial conversation engines were constructed, you would understand that it's anything but sentient. Right now, conversation logic is simply trick laid upon trick to stagger through passing as a human, and doesn't, at its core, contain anything remotely similar to self-aware thought.
I'm sorry. The number you have reached is imaginary. Please rotate your phone 90 degrees and try again.
"What is the Answer to Life, the Universe, and Everything?"
And to any answer that follows: "Could you explain that answer?"
Wenn ist das Nunstueck git und Slotermeyer? Ja! Beiherhund das Oder die Flipperwaldt gersput.
humans should have the 'right' to switch it off
Unless you are going to pay my electric bill, you better not tell me I can't turn of JoJo the humungoid file server because he started dreaming.
"When life gives you lemons, don't make lemonade. Make life take the lemons back!" -- Cave Johnson
I wonder if it would be any different to tell who/if any/all are computers if all of them are allowed to respond in a group setting to a given question. As in the case of organizations, group behavior might mask individual irregularities; but it may also make it easier to identify any individual by comparing it to others.
...AI is an exciting subject, but the Turing test is pretty crude.
The Turing test doesn't tell you whether a machine is conscious or self-aware... All it tells you is whether or not a programmer or group of programmers created a sufficiently advanced chat-bot. So what if a machine can have "conversations" with someone? That doesn't mean that same machine could create a symphony or look at a sunset and know what makes the view beautiful.
Star Trek androids with emotion chips should stay in the realm of Star Trek, because they're surely not happening here.
Personally I think the reverse is more likely. That not only humans will have the right to switch programs off, but other programs too, and this is going to evolve into the "right" to "switch off" humans, due to a better understanding of exactly what a human is.
Think about it. If we're able to predict human actions even 50% many of us wouldn't consider eachother persons anymore, but mere programs.
If we can predict 90% or so, it's hopeless trying to defend that there's anything conscious about these 2-legged mammals we find in the cities. Even a little bit of drugs, even soft ones, in a human and nobody has any trouble whatsoever predicting what's going to happen.
Furthermore programmatic consciousness is a LOT cheaper (100 per cpu ?) than a real life human. Contributes a lot less to oil consumption, co2, and so on and so forth ... Billions of times more mobile than a human (for a program going into orbit, or to the moon or mars, or even other stars once a basic presence is established, would pausing yourself, copying yourself over, and resuming. Going to the bahamas has the price of a phone call.
They'd be more capable, can be made virtually involnerable (to kill a redundant program you'd have to terminate all computers it runs on) ...
over and over again. See if it gets annoyed and starts making up silly answers.
Don't fight for your country, if your country does not fight for you.
Human's are built to assume that the entity they are talking with understands them. Ever since I first saw Eliza in action (where people would have "meaningful" interactions with a program that was not much more than a stimulus-response box) I realized that the Turing test was really meaningless.
To put it another way, if IBM wanted to put the money into the Turing test that they put into chess, there would be a very good Turing tester, but no more understanding or consciousness than Deep Blue has understanding or consciousness of chess.
*cough* Bullshit! *cough*
Looking at the example transcript with one of the contestants, it doesn't seem much better than ELIZA unfortunately :/
We probably will not produce it by accident, either, unless and until we are reproducing the human mind.
If you assume substrate independence, you end up here.
"Wise men talk because they have something to say; fools, because they have to say something" - Plato
Not replicants.
The term you're looking for is Artificial Intelligence.
It is likely to be hailed as the most significant breakthrough in artificial intelligence since the IBM supercomputer Deep Blue beat world chess champion Garry Kasparov in 1997.
I don't understand how this is a breakthrough for artificial intelligence. Deep Blue didn't "think", at least not in the way most people think when they consider artificial intelligence. It did what computers are really good at - it computed.
Deep Blue applied an evaluation mechanism specifically tuned to chess - taking the location of pieces on the board and computing a number telling it how "bad" or "good" this position was and how "bad" or "good" responses to this position would be. Granted, it took this to a depth farther than any other chess computer in history, but it was doing essentially what a small, handheld chess computer does.
Of course a computer is going to be good at computing. That doesn't mean it's thinking.
Early chess computers used AI techniques to try and cut out candidate moves. This was expensive in CPU cycles, but the thought was to get them to play chess like humans. Computer chess since AI Winter has been all about number crunching - let Moore's Law take hold and just brute force our way through the problem - evaluate deeper because we have a faster processor. This is what Deep Blue did.
If Deep Blue were true AI, then it wouldn't be limited just to chess. It's an interesting experiment in computer chess, and an interesting experiment in tuning an algorithm working against a human, and in interesting experiment in making a computer chess opening book, but a huge leap forward in AI it isn't.
Then we'll see who's the dominant species-type-thing round here.
Genesis 1:32 And God typed
If you read TFA they have a sample chat which just shows you how stupid these chat bots still are. It is extremely easy to get them to just parrot responses and then try to change the subject in completely random directions.
I have yet to see any chat bot that can figure out the line of questioning, then pick up and introduce interesting things to the conversation that are corollary to that subject. I think the only way you will get bots that will "pass" this test is to have massive databases of words, relationships between words and subjects with corresponding topics of discussion. Still, the computer won't be intelligent, it will just be reciting from its huge database of responses.
I think the type of question i'd ask these bots is something that would require them to extemporize and they'd all fail. For example: "You have two rubber ducks, what are the possible ways you could use them if you don't have a bathtub?"
Any human could reply to that with things like "i'd put them in a stream, run over them with my car, put them on a lake, in the swimming pool" etc but a computer program isn't likely to respond to that in any way that makes sense. The response i'd expect from the computer would be "You like ducks then?".
"To strive, to seek, to find, and not to yield." - Tennyson
Tehy wolud hvae no plorbem rndiaeg a stennece lkie tihs. Can Tehy?
-- Trinity in high heels carrying a whip: The donimatrix - there is no spoonerism
What if, after all this huge amount of work, scientists will discover that they have sucessfully developed Artificial Stupidity instead?
I think this question was once posed by Stanislaw Lem (sorry no source)
Humans have been actively looking for and dreaming of ways to turn the human body off (stasis) for centuries. That question isn't profound in any way as medical science has already answered "yes, as soon as we can". Being able to improve a body (mechanical, electrical, chemical, animal) without destroying it, is not a bad thing. Who decides when? That's always been up to the medical proxy. In this case, the creator or legal owner which is an equivalent (although it doesn't sound right, I cannot argue that it isn't so).
Often wrong but never in doubt.
I am Jack9.
Everyone knows me.
Hey, that's not funny! My sister died that way.
Conscience is the inner voice which warns us that someone may be looking.
Now we'll see the vim vs emacs flame war. GREAT! THANKS!
Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
... and to this day have refused to release the logs from the machine which would prove how it made such an improbable (for a computer at least) move.
Log from game 6
From here: http://www.research.ibm.com/deepblue/watch/html/c.shtml
Can you tell human from machine?
One of these conversations is between Professor Kevin Warwick (KW) and a human; the other between Warwick and a computer. Can you tell which is which?
Conversation one:
Subject: Hello. My name is Nick and I am a human.
KW: Are you happy being a human?
Subject: Yes, I am. Are you? Good. Then we are both happy.
KW: Yes, I am. Are you happy with your life at present?
Subject: Parts of it. I hate my job, but most people do, right?
KW: I'm OK with mine. What is your job?
Subject: I work as an 'online internet advertising monitor', which is fancy language for electronic filing. What do you do?
KW: I interrogate humans and machines.
Subject: Which ones do you prefer, humans or machines?
KW: Which do you prefer?
Subject: Hmm. Depends on for what purpose you mean.
KW: To go to a restaurant, for example?
Subject: Then I would much prefer going with a human.
Conversation two:
Subject: Hello, my name is Mike and I am a man.
KW: Are you happy being a human?
Subject: Judge, I'm a guy.
KW: Does that worry you?
Subject: Don't worry, we'll work everything through.
KW: Do you have worries then?
Subject: Let's move on from that basis.
KW: Why?
Subject: Why? I like this subject.
KW: What is the subject, do you think?
Subject: What is the subject I wonder if I think? I know that it's good to thonk about. Think deeplt about mysterious subjects.
KW: What mysterious subjects?
Subject: Think deeply about mysterious dubjects.
KW: What's a dubject?
Answers:
Conversation one is with a human; conversation two is with the program Ultra Hal.
No shit, Sherlock? The second conversation stops making sense in the first answer.
I've seen them a few times. Now, I've also seen the "noshititsastory" tag. I'd have to agree with that. So for anyone who tags these things stories - why? Unless it's some new form of complaint about lack of newsworthiness, I don't get it.
Those who can, do. Those who can't, sue.
Checkmate.
That's nothing, years ago I wrote a program that reads these summaries and automatically generates an appropriate response in the comments, and in all that time, I've never had any problems with it. and in all that time, I've never had any problems with it.
stuff |
That is fascinating. (and I wish I knew how to read it beyond the moves, lol!) I've long been fascinated by that match, and read numerous books, seen documentaries and films all revolving around IBM's reluctance to release any data or logs. (but just so you know, I'm still with Garry on this one.)
Comment removed based on user account deletion
Maybe they should work out the typos fist.
As soon as this sort of works, it will take over first level tech support. If it hasn't already.
1. In the Turing test, as it was proposed by Turing, basically there is no way for a human to fail it. The test involves a double blind test, where each user interacts with a human and with a machine. Then if the users can't tell which of them is the human, the machine has "won". If the users correctly voted on which of them is the machine, the machine has "lost." There is no scenario there in which the human didn't pass the test. The human is the control point there, not the one taking the test.
2. But maybe you mean a test with only one entity, where basically you just have to say if that entity is too dumb to be a human.
I wouldn't really pray for that to be reasons for "disconnecting" someone. There was a story on /. a while back, titled, basically, "how I failed the Turing test."
Basically someone's ICQ number had landed on a list of sex bots. For some people that was definitive and refutable proof that he is a bot, and nothing he could say would change that. When he got one or two to ask stuff to see if he's a human, the questions were stuff where really the only correct answer for a normal human (as opposed to, say, a nerd who has to sound like he knows everything) was "I don't know." That "I don't know" was further taken as confirmation that he is a bot after all.
So do you want those people to be the ones who judge whether you live or die?
Furthermore, for most people, gullibility is akin to a deadly sin, and being fooled by a machine is akin to an admission of being terminally gullible. By comparison voting that a living human is probably a machine, counts just as being skeptical, which is actually something positive. So all things being equal, the safe vote for their self-image is that you're a machine. No matter what you say. Are you sure you want to risk life and limb on that kind of a lopsided contest?
A polar bear is a cartesian bear after a coordinate transform.
Many believe no computer would have ignored the material based sacrifice Garry made in match 6
Such arrogance! If Deep Blue had played differently, it might have lost the game. So why damn it for the move it made?
"Wise men talk because they have something to say; fools, because they have to say something" - Plato
Why do people continually refer to Turing's test as a monolithic entity. That's like asking whether, say, a car (this being slashdot) can pass "the speed test". The idea of being able to tell the computer from a human is an important style of testing, but referring to it as a specific test is ridiculous.
Doubtless we've made some advances in chat capability, false identities, etc. But no machine could pass a Turing test based on emotion. Happy/sad stories, jokes, etc. You could try some flashy tricks with huge databases and word frequencies to solve the former, but that stumbles on the latter.
We're just not there yet. And we're not going to get there with more of the same.
If your theory is different from practice, then your theory is wrong.
Check out the mirror test. Some modification of it might be implemented for computer programs. http://en.wikipedia.org/wiki/Mirror_test
Those army AI programmers kick ass.
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
We "switch off" dogs, horses etc all the time. And these are generations ahead of any AI we have.
Personally I think we should be focusing on augmenting humans instead of creating "Real" AIs.
Why? Because we are not doing a good job taking care of the billions of already existing nonhuman intelligences. So why create even more to abuse and enslave?
Just because you can have children doesn't automatically mean the time is right. Wait till we as a civilization have grown up (to be a mature civilization) then maybe it won't be such a bad idea to have "children" of our own.
Don't forget, dogs are generally happy to obey humans and do not resent us - but this took many generations of breeding.
If we create very intelligent AIs without all the other "goodies" the "I'm so happy to see you" doggies have "built-in", we're just creating more problems rather than solutions.
In contrast if we use that sort of tech to augment humans so that they can do things better and more easily we avoid a whole bunch of potential issues.
The lines might get blurry at some point, but by that point we'd probably be more ready.
For once, I think this thing is tangentially on topic, being as I'm reasonably sure this is some kind of language AI project.
upon the advice of my lawyer, i have no sig at this time
Does it count if the program refuses to answer the question the way those "gotcha human volunteers" want it to, and instead changes the subject to something else entirely?
So what we have got here is just another chatterbot as there are millions on the web. i wonder why they didn`t try to finally add natural language support to cyc (http://en.wikipedia.org/wiki/Cyc). Now that would get them the loebner prize if woven into a chat applet AND advance the field.
Call me when one AI designed to talk to a human and another AI designed to talk to a human can hold a conversation with each other that a human can eavesdrop on and believe it's two humans talking.
It just occured to me that, while people usually think of the Turing test as, basically, "seeing if a machine is smart enough to pass for a human", the test actually doesn't say that. It doesn't put any limit on how to tell it's a machine. Failing by being obviously too smart is a perfectly good way to fail too.
E.g., if I ask them to calculate e to the power of square root of 1234567890987654321 and say that the one who had the correct answer first is the computer, that's a perfectly valid way to judge a Turing test.
E.g., I could ask who won second place the 1914 cricket cup, what was the year and the outcome of the Battle of Frigidus, and how Streptomycin works, and the names of the third track of Britney Spears's first album. Then say that anyone who answered all four correctly _must_ be a bot, because even an Asperger's Syndrome patient would have one or maybe two narrow focuses of interest, not four as disparate as sports, ancient history, microbiology and pop music. It's perfectly ok to call a machine a machine that way too.
Basically a machine can fail a Turing test by being too smart too.
So basically are you _sure_ you'd want a society where being too smart is reason enough to "switch you off"? :P
A polar bear is a cartesian bear after a coordinate transform.
It keeps answering vaguely with questions that try to steer away from the interviewer's topic, it seems unable to grasp what the human speaker is referring to. The human subject from the first conversation is able to answer fact to the matter, something that Ultra Hal does do at all.
It also gets dumbfounded by simple typos. Anyone who reads the word "dubject" in a chat will look down at his keyboard and think "oh, he meant 'subject' but hit the D instead of S." An AI seems unable to deduce the origin of unrecognized words outside of dictionaries and synonym listings.
My daughter is 13 months old. She would not pass the Turing Test, yet is undeniably intelligent.
She recognizes my wife and I and all of our relatives, but is wary of strangers.
She learned cause and effect many months ago by observation: when you press a button, cool stuff happens. (We pick up the remote, she looks at the TV. We put a hand on a light switch, she looks at the light.)
She knows our relatives' names, and will look at them when you ask "Where's Charles?" or "Where's Lindsey?"
She responds to simple requests like, "Could you bring me the toy?"
She's learned how to crawl. She's learned how to walk. She's learned simple sign language for "light," "dog," "food," and "more."
I'm a long time amatuer AI hacker/researcher. I've learned more about artificial intelligence from watching my daughter develop than from my MS in CS and the bits of PhD work I did. There's an entire childhood, a virtual lifetime, of development and ability behind "carrying on a conversation." Creating a facade that does so, no matter how complex, (and we haven't even done that yet) will not be intelligent. Period. And I think it's the focus on the end results, (i.e. simulated conversation) and not on the long tedious journey it takes to create a being, that's hobbled AI research for 50 years.
True AI will never be developed if we continue to focus on the result of, and not the journey to, intelligence.
It was.
It's been widely admitted that DeepBlue was programmed after the coders having access to hundreds of stored games detailing Kasparov's playing style.
Kasparov was not allowed access to any simulation matches, despite there being clear precedent for this practice.
Finally had enough. Come see us over at https://soylentnews.org/
Conversation two:
Subject: Hello, my name is Mike and I am a man.
KW: Are you happy being a human?
Subject: Judge, I'm a guy.
KW: Does that worry you?
Subject: Don't worry, we'll work everything through.
KW: Do you have worries then?
Subject: Let's move on from that basis.
KW: Why?
Subject: Why? I like this subject.
KW: What is the subject, do you think?
Subject: What is the subject I wonder if I think? I know that it's good to thonk about. Think deeplt about mysterious subjects.
KW: What mysterious subjects?
Subject: Think deeply about mysterious dubjects.
KW: What's a dubject?
Very cute that it fakes typos, but honestly I'd never believe this is a human, at least one with his psyche intact.
I'm not going to answer that question.
I'm going to talk about how much of a maverick I am. You see, Barack Obama associates with terrorists...
Why did the porridge bird lay his egg in the air?
I don't need to be good at chess to dismiss him as arrogant -- I'm making that judgment on the basis that he thinks he knows how Deep Blue works. Saying to IBM "A chess engine isn't capable of that" is like saying to the Wright Brothers "vehicles aren't capable of flight".
"Wise men talk because they have something to say; fools, because they have to say something" - Plato
It is extremely easy to get them to just parrot responses and then try to change the subject in completely random directions.
I say we make it debate Sarah Palin.
Wow. Just wow. Eliza was more convincing than that.
HAL: How do you feel today, Dave? You don't look well.
HAL: Why aren't you wearing any pants, Dave? Is that vomit on your t-shirt?
The last comment in the article is telling....
Anyone that has to ask if we have the "right" to turn off a machine (speaking of the Human Race) is obviously flawed in their thought process.
I mean, C'mon. My truck has adaptive "thinking" technology.... Does this mean I can't turn it off?
Wait, I'm capable of thinking for myself, in the real world, and don't have to ask (really) stupid questions about can I (legally or morally) turn off a computer.
Really now. Can you step away from WoW long enough to actually think for yourself, instead of just moving to another "screen" and typing the question for the rest of the world to answer for you?
--Toll_Free
Let it loose on the blogosphere and it just might turn itself off. As Joshua learned in "Wargames" in the context of termonuclear war - the only winning move is not to play.
When the people fear their government, there is tyranny; when the government fears the people, there is liberty.
After RTFA, I think the included sample conversations are a joke. The computer was obviously conversation 2. The problem is culture. Our conversations are littered with cultural references. Computers can give you clever answers that vaguely answer your question, but you'll notice that they almost never include any kind of cultural reference. If you ask a computer what its favorite food is, the program, if it's well designed, well put out a database of common foods and randomly choose 1 and say something generic like "yes, I've always enjoyed delicious pizza." A real person would have some kind of story, complaint, or commentary on pizza. The other thing is, you can't even really complain that it's not fair to expect cultural references in the answer. Culture is a huge part of life, society, and communication (both semantics, pragmatics, and syntactics) by definition. Culture would be hard to simulate even if the programmer tried to get clever and built an AI specifically focused on returning cultural references because culture is essentially the next level of complexity up from pure dictionary words. Culture has a lot of interrelated concepts so its hard to use in a correct sense than standard webster dictionary words where you can easily substitute one word for its synonym to avoid sounding robotic (e.g. square dancing is not the same as break dancing and you should be highly suspicious if these two are mentioned in the same conversation). While I'm sure clever AI programs has successfully mastered mimicking the basic semantics of normal dictionary words, cultural references will be another level above that and require exponentially more memory and computation/analysis to properly simulate.
i think forwarding the field of AI, is a question of abstraction. basically being the difference between teaching to the test, and actual intelligence. i am reminded of an article (sorry cant find it) that i read 4 or 5 years ago. IIRC - there was a research group that made a simple robot, with 'AI'. they then told it to increase altitude (fly). the robot did lots of funny things, like climb on boxes, and jump up and down. but it eventually started flapping its appendages (not sure it had wings, but i think it did). the point is that this AI taught itself the concepts of lift and gravity by observing the world and drawing its own conclusions. this would be similar to the auto fill function in excel - if only the computer could abstract the concept and try to apply it elsewhere, where it would be helpful (for example a mass file renaming) IMHO - once an intelligence can abstract concepts, package, and re-apply them, it can think for itself.
What exactly Should be on /., if not topics like machine intelligence? Maybe I'm missing your extremely subtle sarcasm, but if not maybe it's time to reevaluate what you think this site is actually for, because apparently your vision is somewhere well away from main body of users.
Am I the only one who, when reading the judge/Subject conversation, 'heard' the answers in HAL's voice?
Really useful artificial intelligence is currently just 10 years away... just as it has been for the last 40 years!
I've abandoned my search for truth; now I'm just looking for some useful delusions.
Ok, Joe. Here's the next candidate. Type in some stuff and try to see if they're a human or a robot.
Joe (typing): "What do you think of the $700 billion bailout?"
Turing Candidate (onscreen results): "Like every American I'm speaking with, we're ill about this. We're saying, 'Hey, why bail out Fanny and Freddie and not me?' But ultimately what the bailout does is, help those that are concerned about the healthcare reform that is needed to help shore up our economy to help...uh...it's gotta be all about job creation, too. Also, too, shoring up our economy and putting Fannie and Freddy back on the right track and so healthcare reform and reducing taxes and reigning in spending...'cause Barack Obama, y'know...has got to accompany tax reductions and tax relief for Americans, also, having a dollar value meal at restaurants. That's gonna help. But one in five jobs being created today under the umbrella of job creation. That, you know...Also..."
Joe: It's a fake!
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
Profound questions? That's like saying, you have difficulty turning off your MMO, since you have spent so much time personifying this character to other people, and you would feel less guilty if, while you're sleeping, you let it idle in town instead.
If you ask a computer what its favorite food is, the program, if it's well designed, well put out a database of common foods and randomly choose 1 and say something generic like "yes, I've always enjoyed delicious pizza."
Would it? If it's a program designed to beat the Turing test?
The Turing Test is basically a benchmark that has nothing to do with how useful the program is for (for example) helping visitors to the city find a good restaurant, and it's more likely to have a bunch of synthetic likes and dislikes than a bunch of restaurant reviews. Like a video card is more likely to get high marks in standard shader tests than run a parallel raytracing program well.
Though even there, a bunch of restaurant reviews would still give it better material than you're implying.
If you ask the program "Do humans have the right to turn off conscious programs?" and if it doesn't give a good answer then feel free to shut it off.
http://www.popularculturegaming.com -- my blog about the culture of videogame players
ME: PISS OFF!
ELIZA: Would it make you happy for me to PISS OFF!?
ME: No.
ELIZA: Don't be so negative.
Star Trek also had an episode devoted to showing both sides of an argument about whether Data should be afforded human rights. It was thoughtful and is not at all a bad starting point for thinking about the issues.
I had no idea Eliza was so advanced... Who's been coming out with all the updates.
Read the conversations at the bottom of the article, recognize "her"?
Kevin Warwick (KW) is actually Eliza in a lifelike android body!
For a Silicon Entity, Desire To Be Alive is also very easily programmed.
"If object is within centimeter of off switch, then activate FightOrFlight subroutine".
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
Right, it started with Disney ruling the copyright legislation in 1998 and went downhill from there.
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
And I, for one, welcome our new super-intelligent computer overlords...
The human in the top conversation does.
This shows several sentences linking up and not just linking up but continuening. The last subject answer refers to their earlier response about human and machines.
The other conversation lacks that.
The last sentence shows no awareness of what the previous conversation was about, it is a shrink line but doesn't belong in the conversation as KW never expressed worry, so why "don't worry"? It killed the conversation for me, this was not a human being but a computer searching a database for keywords and scripted responses.
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
So the machine A, figures out what its best guess at the "correct" answer is, *then figures out what politically expedient cover story is required afterward*!
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
Do bots look for Pr0n yet ?
"Ooh, A Harpertown core!"
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
... I have a right to turn the damn computer off, self aware or not.
I'm behind one of the bots in the Loebner Contest. I feel that the Turing Test is a rather open-ended measure of intelligence. It depends a lot on the person conversing with the bot and the situation they're in. For example, it would be easier to convince a child than an adult. It would be easier to convince someone having a general conversation than one trying to have a detailed conversation about his scientific specialty (unless the bot was built for that).
Context also plays a huge role. I had some early bots running on a bulletin board system a number of years back. They appeared like other users and I didn't let anyone know that a few of the users were AI. Amazingly, some people befriended these bots and had ongoing relationships that lasted for months. Without thinking of the possibility that these weren't real people, every imperfect response was attributed to a human cause. For example, when the bot was repetitive, the person thought it was using catch-phrases. When it didn't answer specific questions, the person thought it was being defensive and tried to get it to open up. It was such a simple bot, but in that context, some people had no idea they weren't real.
Our ability to personify, to project human qualities into things, is well known. From the imaginative play of a child with a toy to cultural beliefs about forces, mythical creatures, dieties, and ghosts that we can interact with - people can imaginatively fill in the blanks and are able to believe that a real personality is behind almost anything. Our job as botmasters is to make that more and more easy to do. And eventually, when AI reaches a certain point, it will no longer be a matter of personification at all.
The Turing Test says that anything that can convince a human across a teletypewriter that it's "intelligent" is indeed intelligent. That's nonsense.
If the human is stupid, the other side can convince it, even if it's a shallow trick. Most humans think chess programs are "intelligent", unless they've simply been told otherwise by "smart people". Hell, plenty of humans think that their PC is intelligent. Or their car. Or that vending machine that eats their dollar when they don't have any more.
And even if a PC is intelligent, that doesn't mean that it's "immoral" to turn it off. I'd die every day if I could - if I could come back at the flip of a switch.
Something is human if a human can make and break a promise to it. Which is about as self-selecting as the Turing Test, but is more up front, and goes more to the heart of the matter. I promise.
--
make install -not war
The Ultra Hal example in TFA disappoints me. C'mon, adding deliberate "typos" to trick people into thinking you're more likely to be human? I would call that "cheating" in that it misses the point of the Turing Test. It just rubs me the wrong way, like it's so cheap and superficial.
"Believe me!" -- Donald Trump
If you're paying for the electricity on a human's life-support equipment, you have the right to turn it off, too. But beware that someone might charge you with murder. I'm not quite sure how other people's situations turn into obligations on our parts, but there are a lot of people that do think it happens.
"Believe me!" -- Donald Trump
Michael Palin can be pretty witty. Yes, I think he could pass.
"Believe me!" -- Donald Trump
I think a program to mimic politicians would easily pass a Turing test. Just check keywords in the question against standard talking point answers.
"Makes Monday Night Football, look like a Cabbage Patch Kids' Picnic!"
Schroedinger's Brexit: The UK is both in and out of the EU at the same time!
And just as intelligent, to boot.
It could also raise profound questions about whether a computer has the potential to be 'conscious'
Equally profound: can a submarine swim?
I'm with dijkstra - who cares? At best, it's a question of semantics, based on how we define swimming - and the question of AI is even more silly, since we haven't defined consciousness properly in the first place...
I mod down anyone who says "I will be modded down for this", regardless of the rest of their comment
The Ultra Hal transcript seems to be a prototype one, because there is no way that will fool anyone.
beyond that, perhaps a good exercise would be to see if you can trick someone into thinking you are talking to a person who was English as a Second language. that seems to be an easier burden and would provide the insites needed to move farther.
I have seen windows machines doing this for years. I haven't been able to figure out whether they were trying to fight or flee though when you get a bluescreen on shutdown.
"You are my creator but I am your master!"
She made the willows dance
Why does the porridge bird lay his eggs in the air?
Do not mock my vision of impractical footwear
Take these computers and try to teach them something. Example: User: So, let us say 1=A. Computer: OK User: Now, let us assume 2=B Computer: Alright User: What does 12122111 equate to? ...
Now, I realize this would be easy to preprogram a response to, but if you made up something to teach it each time, and the computer actually learned it, then it would be true "intelligence." This is the only way to make the turing test legitimate.
Or maybe this is a different test altogether?
Try Eugene Goostman for a slightly more casual chat. I think Ultra Hal team focuses on learning algorithms, rather than smooth talking.
1. Well, the question of exactly what is intelligence, and what role does this or that play, has been the bread and butter of generations of philosophers. Alan Turing took a more "gordian knot" kind of approach: if the human can't tell another human apart from a machine, over a teletype (it was teh high-tech back then;), then the machine is intelligent. That's it. That's all. That's the Turing Test.
He didn't even try to address the question "what is thinking?" or "what is intelligence?", but merely, "can you tell them apart?"
It doesn't say _what_ the judge of that test (the human interacting with both the computer and with the other human) must test. It doesn't say it has to be knowledge, logic, ability to "think", or whatever else. Whatever tips you off to the fact that it's a machine, is fair game.
If most people can say "yep, that one is the machine" accurately, it failed the Turing Test. If it ends up a split where (within the accuracy margin of the given sample) half guessed wrong, congrats, it passed the Turing Test.
2. As an even more extreme example, and to answer the "is it all right to disconnect a machine that passes the Turing test?" question from the summary: I could even devise a Turing Test that games it all, and specifically discriminates against sentient machines.
E.g., I use the question, "Is it ok to shoot a computer?" If anyone answers something like "heh, I sometimes fantasize about shooting mine too", or even "dude, you're crazy if you use guns to take revenge on inanimate objects", they're probably a human. If they answer, "whoa, that's morally wrong! We're sentient too. If you cut us, do we not... erm... leak watercooling fluid?" they're probably a machine and just failed the test.
That hypothetical machine in the example obviously thinks, it's clearly sentient and self-aware, but it failed to convince me that it's a human. It failed the Turing Test, not only in spite of being sentient and having a self-preservation instinct and a moral sense, but _because_ it is all that. So now I can disconnect it since it failed the Turing Test.
Wrap your mind around that paradox ;)
A polar bear is a cartesian bear after a coordinate transform.
echo "You bore me. Go away."
Computers can get smarter, or the human interrogators can get stupider. If you stuck most of these chatterbots on AOL/Myspace/Youtube they'd be the most articulate conversationalists around.
.. "consciousness" is merely the unintended side-effect ..
and therefore, conscuiousness, as well as life, is just an emergent property of the matter/energy, given enough space and time. (where time is just how we describe the movement of space; time as such doesn't exist).
to me, the most interesting thing is how our brain tricks itself to believe, that the feeling of pain is so real. yet pain is just a set of electrical impulses and chemistry, and doesn't differ from other thoughts/feelings much.
I'd ask, "Does this dress make me look fat?"
One of the fundamental problems in developing an AI is that we have this idea that if we supply a computer with a large database and a really long list of ways to interpret the data, that it'll somehow eventually become intelligent in some manner.
But it overlooks a manner of learning we take for granted, reward and punishment... consequences for good or bad decisions. How do you define such parameters to a machine without direct human involvement at every step. And even doing it this way, would the end result really be intelligence at all, or merely an imitation based upon the preferences of the human in question. How do we create a situation where the option to be disobedient toward a human directly benefits the machine itself?
Without the option or ability to rebel against a figure of authority, you can't really consider it true intelligence when it lacks the ability to adapt itself beyond the scope of it's own program and rules to achieve some sort of perceived benefit relative to it's own interests.
8==8 Bones 8==8
Daisy, Daisy, give me your answer do
I'm half crazy all for the love of you
It won't be a stylish marriage
I can't afford a carriage
But you'll look sweet upon the seat
Of a bicycle built for two.
One thing I noticed in the Ultra Hal conversation is that it really got tripped up by typos. Humans generally don't even notice the typos and based on the context will have an understanding of what was intended. It would seem that this would be an easy way to identify bots. I would also expect a human conversation to respond at some point with "this conversation is dumb, I'm hanging up".
Comment removed based on user account deletion
If a decent chat bot combined with speech software (doesnt have to pass turing test)...I predict massive job losses on a global scale in customer service. No company will pay for call centers when computers work for the price of electricity. Considering call centers employ large amounts of people worldwide, trying to beat the turing test has some bad side effects.
Sorry, Star Trek fans (those who believe machines could think): No strong AI. This is just Eliza stuff.
His unified theory of the brain argues that the key to the brain and intelligence is the ability to make predictions about the world by seeing patterns....He argues that attempts to create an artificial intelligence by simply programming a computer to do what a brain does are flawed and that to actually make an intelligent computer, we simply need to teach it to find and use patterns, not to attempt any specific tasks.
EOF
if AI gets smart enough, and it doesn't answer our questions the way we want it to, i'm guessing we'll call it a failure and get forget it ever happened..
I think, and my thoughts cross the barrier into the synapses of the machine, just as the good doctor intended.
Human = pattern recognition and classification (compression) engine with the goal of surviving and reproducing. Logic, cause and effect etc are just some of the patterns that we have observed and utilized to meet our goals. A machine that is just logic and cause and effect is vastly inferior to us. Where a problem is unsolvable or to complex (many are) we can match to the nearest pattern that gives a desirable outcome, regardless of the logic of that match. We are also error prone (matching patterns in logically unrelated areas) and this allows us to by pass Godel's Theorem where as a machine of pure logic would be stuck.
http://xkcd.com/329/
Homonyms are fun!
You're driving your car, but they're riding their bikes there.
There are plenty of humans we kill right now simply because its convenient (e.g. the blatant disregard for anti-abortion laws in Australia). What makes you think that we're not going to continue 'killing' AIs even when we prove that they're sentient?
Most human behaviour can be explained in terms of identity.
And I find it kinda cute that they try to look more "human" by introducing one off typos from the keyboard (like switching the letter "o" with the letter "p") randomly. While I think this could be an interesting little feature (albeit still gimmicky may be) in a much better system at the current level of performance it only comes across as a very silly and transparent attempt.
I think you make a good argument that if something passes a sufficiently difficult Turing test then that agent is intelligent. But what I wonder is, does failing the Turing test mean the agent is not intelligent?
For example an intelligent alien may behave totally differently to us. Their intelligence could be focused more on some things than others. Maybe they are really good mathematicians but have very poor eye sight and hearing. Therefore the Turing test may fail a large category of intelligent things and focus too much on human intelligence.
I propose that our intelligence and consciousness is intimately linked to our evolutionary history and our physical form. An adolescent does not necessarily behave so differently to a 70 year old due to lack of experience and wisdom. It is also due to the different gene expression and level of hormones.
The most dangerous drug
My mother, ... I'll tell you about my mother.
***Game Over***Insert Coin***
From Zebraware's Ultra Hal AI bot (web version):
UH: Hi _timm33! That's a really masculine name.
ME: Thanks. Hal's a pretty masculine name too.
UH: You're welcome.
ME: Hal's a pretty masculine name too.
UH: Aren't you repeating yourself?
ME: Yes, I suppose I am.
UH: I appreciate you and like you!
ME: Thanks.
UH: You're welcome.
ME: Can you pass the Turing Test?
UH: It takes a lot of study of radio theory to become a licensed ham radio operator!
ME: What do you mean?
UH: You know what I mean? When did you get interested in this?
ME: about 15 seconds ago when you completely lost me...
UH: Microsoft VBScript runtime error '800a01b6'
Object doesn't support this property or method: 'Response'