Slashdot Mirror
AI Researchers Say 'Rascals' Might Pass Turing Test
An anonymous reader writes "Passing the Turing test is the holy grail of artificial intelligence (AI) and now researchers claim it may be possible using the world's fastest supercomputer (IBM's Blue Gene). This version of the Turing test pits a human conversing with a synthetic character powered by Rascals software crafted at Rensselaer Polytechnic Institute. RPI is aiming to pass AI's final exam this fall, by pairing the most powerful university-based supercomputing system in the world with its new multimedia group which is designing a holodeck, a la Star Trek."
Will it have a little AIBO dog with a ring around one eye?
...want the history books to report that the FIRST AI was a Rascal?
I didn't read the article, but at first glance thought the title was "racists might pass Turing test."
End the FUD
I think the people behind this misunderstand the difficulty (and purpose) of passing the Turing test. The problem isn't in manufacturing a believable back story for your program's "character". The problem is in communicating effectively in spite of the inherent ambiguity, fuzziness, and confusion of human languages. I think it's very unlikely that any team is about to meet this threshold.
The most rabid believers in American Exceptionalism are the exact same people whose policies are destroying it.
You're right! They should call it "artificial intelligence" or something like that.
Well, it ISN'T skynet - the authors made an AI character that they could talk with - that wouldn't mind if they totally geek out on WOW topics or file system discussions ;)
I'm not even sure it is actually passing the full Turing test. From the article I noticed one of the researchers saying: "That's how we plan to pass this limited version of the Turing test." Anyone know what he means by this being a "limited" version of the Turing test?
That makes this approach all the more interesting.
Don't take my posts literally; it's just code to control my botnet.
The Turing Test for AI says that if an AI can fool a human into thinking it's human by communicating over a teletype, then it's really "intelligent".
:P.
That's hogwash. Any number of real people I talk to could easily be simulated by some non-intelligent machine. Especially over the phone, to tech support etc.
Slashdot alone is proof of the fallacy of the Turing Test. Unless all you ACs and TrollMods are actually bots. Or maybe it's me. That would explain a lot
--
make install -not war
It is interesting that they have used a 'guinea pig' student to 'bare all' to the knowledge base. It would seem, then that this AI is in fact a type of facsimile of this student.
As we become more comfortable with accepting communication with each other through more abstracted proxies - like common chat applications currently and the recent neural voice collar (which pumps out a synthetic voice - even further proxy) - I wonder if we will in fact see what the author Stephen Baxter speculated, artificial clones of ourselves or our personalities handling our daily affairs.
I don't think it's too far out there to imagine interacting and planning a meeting with someone over the phone, only to find out later you had been talking to an AI facsimile of that individual.
What would (and may) be stranger yet, is considering the possibility that two AI facsimiles may in fact carry out real work or meetings from start to finish completely without the interaction of their 'owners'.
Read my Very Short "Stories"
If they succeed I'll never get my kids out of the basement!
Clearly, in order to pass the Turning test, the AI must be an ambiturner. The easiest way to determine this is to have it turn right, and then ask it to turn left. If it can't do this, it fails, just like Zonk fails at editing.
If the avatar is limited to talking about themselves, their mental state and the mental state of others, it doesn't seem like a true Turing Test. I mean, would a question about flipping a tortoise on its back be allowed?
On a different note, don't they know that giving it "memories" doesn't mean it will pass the Voight-Kampff test?
Reading code is like reading the dictionary - you have to read half of it before you can go back and understand it.
If ever an article needed a "whatcouldpossiblygowrong" tag. Turing Test AI's combined with holodecks? All we need now is to pair it with those carnivore hunter-seeker robots that power themselves with fermented slug-flesh and just wait for them to figure out humans have more meat.
Kwisatz Haderach
Sell the spice to CHOAM
This Mahdi took Shaddam's Throne
just because it can pass the turing test does not mean the machine demonstrates real intelligence!
But it will demonstrate that past a certain point we won't know the difference between real intelligence and something attempting to appear intelligent.
in fact, just what is intelligence / conciousness? if we can't define it, how can we hope to produce it?
If we can't tell the difference maybe there isn't one. Are you intelligent? Or are you just sufficiently complex enough that you simulate it well?
With all due respect to Turing, and he was a brilliant man, I don't think that his test is the definition of AI. True AI must be self programming. Here is the ArcherB test. When you can place a machine in a particular situation with no programming whatsoever, and it figures it does something on its own, then you have AI. For example, hook up a computer to an Internet connection. This computer can have no BIOS, OS, no programming at all. When it learns to use its own hardware, figures out network protocols and starts downloading web pages and porn, you have true AI.
There is no "I disagree" mod for a reason. Flamebait, Troll, and Overrated are not substitutes.
Limiting the topic: In order to limit the amount of area that the contestant programs must be able to cope with, the topic of the conversation was to be strictly limited, both for the contestants and the confederates. The judges were required to stay on the subject in their conversations with the agents.
Limiting the tenor: Further, only behavior evinced during the course of a natural conversation on the single specified topic would be required to be duplicated faithfully by the contestants. The operative rule precluded the use of ``trickery or guile. Judges should respond naturally, as they would in a conversation with another person.'' (The method of choosing judges served as a further measure against excessive judicial sophistication.)
"That's how we plan to pass this limited version of the Turing test."
If it's a limited version of the Turing Test, then it's not the Turing Test. They don't actually define exactly what the limits are. But any open ended test is doomed to failure based on our state of the art in A.I. (read: there is no science of Artificial Intelligence, in the sense of artificial cognition).
"What do you think a typical mother would say if she found out her daughter was going to enter the porn industry."
"Why do you think children have emotional attachments to their parents?"
"Which is worse, racism or sexism?"
"Would you rather be a fireman or an astronaut, and why?"
Any sort of open-ended question that requires human cultural knowledge and asking it to support its conclusion is going to cause it to barf.
Now, if the point of this is whether you can fool someone into thinking the Avatar was human when they didn't know it was a test, well, who cares? Eliza was able to do that back in the 1970s.
Lastly, who says the Turing Test (or any A.I. test) needs to take place in real time? I would be impressed if they came back with a human-level answer in a month of processing time. That's equivalent to a computer 2.5 million times faster than a computer that could produce the answer in one second. That they can't even do that should tell people that speed is not the problem in A.I. research. We have absolutely no fundamental model of how it all works.
Sometimes it's best to just let stupid people be stupid.
For heaven's sake - build a freakin killswitch into the thing!
DISCLAIMER: This post was not checked for speling and grammar- if you complain- you're a whiner
From the summary this "test" is not a strict Turing Test as it appears to be the machine talking to a human, alone, with no second human also talking to the first human. I could be wrong of course.
One of the things that makes this test so special, is that if you cannot tell the difference between a human and a computer, then essentially the computer is intelligent. Why? Because if you cannot tell the difference, what does it matter if the machine is really intelligent or not? Is the machine was really thinking or was it just cleverly programmed? The point is however, if you can't tell the difference, what does it matter? (Incidentally, I apply the same argument to the "question" of "free will".)
Anyway, if this machine (or personality) consistently passes a proper Turing Test, then yeah, that's pretty cool, and I want one on my computer, well so long as the personality type is compatible with my own (not a Marvin please...). (And I have a partner, so no need to make such jokes...)
I wank in the shower.
Besides these two flaws http://en.wikipedia.org/wiki/Turing_test#Weaknesses_of_the_test, I'd suggest the most fundamental flaw is assuming humans are intelligent. I've met a few that you'd have trouble distinguishing from a potato. Doesn't mean the tuber is intelligent.
The world is made by those who show up for the job.
One of the problems for any entity trying to communicate like a human is that we share some common knowledge which is based on our physical existence (pigs can't fly, but fall etc.) Some AI projects like (Open)Cyc have tried to feed their AI with a very large number of simple facts, but to "understand" some concepts you have to experience them. Try to explain the difference between red and blue to someone who was born blind.
The 3D communication (holodeck) aspect mentioned is therefore an attempt to have an AI "living" in a human like space, to enable it to develop a similar world view. What's new about Rascals (Rensselaer Advanced Synthetic Architecture for Living Systems) seems to be something else ("Rascals is based on a core theorem proving engine that deduces results (proves theorems) about the world after pattern-matching its current situation against its knowledge base.") that is very computing intensive. Whether this will make any real difference remains to be seen, a lot of other approaches have failed and they so far have only succeeded with very limited models.
memomo: free web based language trainer DE-EN-ES-FR-IT
A real Turing Test is supposed to be a normal conversation between a human being and a machine, NOT a contrived scenario limited to particular subjects or format.
Visual memory hasn't yet been developed for the computers to use generalizations. Specific real data isn't available to them. Google is trying to use wetware to sort images, process dead links, and form new commerce content. When all three are done completely by computers then they will have enough smarts to pass the turing test reliably.
Curiosity was framed, Ignorance killed the cat.
But can it do THAT:
http://xkcd.com/329/ ?
Don't take my posts literally; it's just code to control my botnet.
>|<*:=
And as I understand it, the test they propose is not only one-sided, it is limited in scope. That is not a Turing Test, in which one is supposed to engage in free conversation. There is a world of difference.
An interesting point.
I suppose what we can do is produce something which carries out tasks which we consider intelligence necessary for - in that case does it really matter if it is intelligence, so long as the 'task' gets completed?
Be that task mathematics, logistics or writing smooth jazz.
I guess perhaps the problem has been that we've been looking for human-like intelligence for these tasks, when really we should be asking what does intelligence do. Instead of asking what intelligence is and how to make it, perhaps we should just be searching for ways to accomplish the tasks intelligence tackles so well.
During the early days of powered flight many found it difficult to give up the notion of flapping wings...after all, since everything that flew under it's own power used wings which flapped, flapping must be needed as well as wings. Rocketry might be an example of flying without wings or flapping.
I guess we can think something along these lines - it doesn't have to flap it's wings to fly.
Read my Very Short "Stories"
consciousness is an illusion humans hold to make us believe we are better.
The Kruger Dunning explains most post on
...the Turing test will be limited to controlling avatars in a virtual world--probably Second Life. Both the synthetic character and his human doppelganger will be operating different avatars. If the human-operators can't tell who the RPI synthetic character is, then it passes the Turing test...
Seriously? Their turing test is on an online game?
This isn't a reasonable test. The way people converse online is MUCH different from how they converse directly. I suspect most of the users on a game like that would fail the turing test. "LOLWTFBBQ! oh hai! pwned!"
Most of the students at RPI don't have enough personality to pass the Turing test. I am doubtful if they could design a program to beat the test.
Virginia is for lovers. EVE is for griefers.
That's why I like to get rid of it as often as possible.
Read my Very Short "Stories"
What is the difference between the Turing Test and the Turing Machine? I thought it was less about mimicking a human and more about generalities that can cause a machine to mimic everything possible thing possible via programming. I'm off to read....
As for a computer, you give it the necessities for life... power and cooling. Let it figure the rest out. I guess I'll give a little and say you can help it along some. Maybe give it a dictionary on the HDD or something and maybe teach it to read. But I'm afraid I'm going to have to stick to the no BIOS, no OS thing. People figure out their hardware on their own. Until a machine can do the same, it will be lacking.
There is no "I disagree" mod for a reason. Flamebait, Troll, and Overrated are not substitutes.
yaaay! design that holodeck for me. the sooner I can move into my virtual world and live my my simulated Monica Bellucci and her three simulated identical sisters the better (though I might have to debug and apply a patch for her personality).
oh wait? they are *only* working on the personality? damn.
oh well, at least the cleaner will not need a mop and bucket.
but seriously: you wait until telemarketers and con men get hold of an artificial personality that can hold several hundred conversations simultaneously - play the numbers and you talk people into anything.
the real turing test is being able to Phish in a chat room. One you can automate that you're golden. and it's pretty unarguable it passed a turing test. Slashdot had a article a while back about robo-chats doing just that but they relied on pretending to be non-native english speakers.
I wonder if it's easier to do this in Japanese than English. From what I've read Japanese is easier to text message in because the object and direct object are usually inferred and there are no cases or articles. A single sentence can be one character and just a verb. Thus by constraining the nuance into discrete choices rather than sparsely populated product space of self-consistent cases, predicates and adjectives, perhaps japanese would be easier to generate turing worthy text.
Or maybe the reverse is true. But I'd bet one was a lot easier than the other.
Some drink at the fountain of knowledge. Others just gargle.
RPI is aiming to pass AI's final exam this fall, by pairing the most powerful university-based supercomputing system in the world with its new multimedia group which is designing a holodeck, a la Star Trek."
Oh wonderful. Now we not only have to talk to troll-bots on the web, now we have to *see* them in person too.
Table-ized A.I.
Just because you can fool a human after interacting for an hour doesn't mean you can keep up the act for a week or a month or a year.
There are lots of computers that can pass a 5-minute version of the test.
No, to really pass this test the computer will have to have and display a definite, self-consistent personality that is consistent over time. It doesn't matter much what this personality is as long as it's self-consistent and credible. A lack of a personality will be picked up on by an observer over time, especially when compared to the real human the observer is also conversing with.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
The inquirer tries to discover if he is talking to a machine or not. Being undetected during casual conversation (and I can bet even this is far far from reaching that, they're just making PR) is one thing, being undetected when tried is different.
Ultimately, the Turing tests tests much more than the ability of conversation. You can describe problems in a conversation and ask the computer to solve them, this is what makes the Turing test a true A.I. test.
\u262D = \u5350
Considering the abysmal performance of the automotive X-Prize competition; the Turing Test might be a little too ambitious for Rascals.
You mean to tell me that the avatars I already interact with in second life are really people? Could of fooled me!
I'd rather have a full bottle in front of me than a full frontal lobotomy.
I'm reading Slashdot => no. QED
Thank God for evolution.
This was the premise of Blade Runner. That's why they developed the Voight-Kampff machine to be able to single out replicants.
Well, you do have to admit that even humans are born with some very basic instincts, such as the desire to suckle when hungry, closing their hand when something is touching their palm, cry when they're uncomfortable (hungry, wet, tired, in pain) as well as the involuntary actions such as cardiopulminary functions.
That said, I would agree that you shouldn't have to give a machine anything more than basic resources to begin its process of learning, but you do need to give it something a rudimentary kernel to get it kick-started from the state of being an inanimate pile of silicon. From that kernel, it should be able to learn from its surroundings, build its own OS and begin to interact with its surroundings.
But there is a genetic basis for the fundamental structure of the brain...
True, we have to essentially figure out how to USE the signals we get from our senses, but the brain already has the basic structure to interpret your senses and do gross movement. (Or did your baby not move it's arms and legs when she was born?)
Therefore, the correct analogy would be the hardware necessary (including BIOS) AND the basic OS. You don't tell your AI how to "read" the internet, but you do tell it how to interpret the signals. So your AI knows that there is something out there and then figures out what it means and starts using it productivly.
Also remember that the "hardware" for the AI could be entierly software based...
You have an excellent point but are taking the analogy too far.
All animals with a central nervous system have some form of BIOS and OS. No one has to 'figure out' how to breath, feed, eliminate waste, or circulate blood. In addition, many behaviors are built in, including the behaviors that let a creature learn more than it was born with. For instance, no baby animals of any sort will voluntarily move off of a cliff, even onto a clear surface that would support them. Fear of heights is built in.
However, I think I see what you are getting at. This is a programmed system, not one that learned most of its behaviors through trial and error. A system that can't start where a baby starts, and can learn the basics on its own the way a baby does, is still lacking. But the "No BIOS, no OS" thing is going a little too far.
- None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
Having spoken at length to Daisy (http://www.leedberg.com/glsoft/daisy/), it seems that after a while a simple AI can be taught enough to become fairly human. Daisy starts with no language, and simply learns a vocabulary from the people with which she speaks. Since the test of "human interaction" is subjective, and Daisy is formed from your conversations, she becomes pretty realistic, and also likeable. On the other hand, people who are nothing like me are more difficult to get along with. Are we not testing "how like me is this?", not "how human is it?"?
Japanese is a pro-drop language, in that you can leave out subjects or objects in speech if it's clear from discourse what you're talking about.
But Japanese definitely has a case system where the inflectional morphology is indicated by particles that follow the modified noun.
Solving the class of NP-complete class of problems is a much loftier goal than making a convincing chatbot.
It's more like the entrance exam. That is, if a computer cannot be reliably distinguished from a human being (within the confines of the test setup), then we MIGHT have something bordering on intelligence. It's a great achievement and a landmark, but it's not the final test.
All I heard was blah blah blah turing test blah blah IBM blah blah designing a holodeck, a la Star Trek .
:)
Passing a turing test is one thing, but a holdeck? Oh yesss....
All the Star Trek officers, engineers, etc. are prudes with their use of the Holodeck. The Ferengi's knew how to sell that product. You know what I'm talking about
Bring it on...
My Turing machine was called "Artificial Insanity". I used to have a copy posted on the internet, but I ran out of room. It was so human that a friend of mine broke his keyboard it pissed him off so much.
I tackled the problem with two ideas: One, humans are stupid, crazy, defensive, argumentative, get drunk, tired, and stoned, and generally behave like... well they generally DON'T behave. Secondly, as it was designed on a Timex-Sinclair 1000 with only 16k of memory and no hard drive, it had to be really, really simple. So I had to resort to trickery to fool people.
One of these days I'm going to port it to javascript and post it.
Once I ran across a Turing machine on the net named "Alice" and had Art have a conversation with it. I think the two machines fell in love with each other! I posted the results at my now-defunct nerdy Quake site, you may still find it at archive.org, even if Google can't.
-mcgrew
mcgrew's razor: Never attribute to stupidity that which can be explained by greedy self-interest
I'd argue our brain and perhaps even our DNA is the equivalent of a BIOS and OS. Humans are even born with certain instincts amounting to preprogrammed instructions, breast-feeding being one of them. A computer with no BIOS or AI is basically a pile of plastic and silicon. There needs to be some foundation to build upon.
The conditions I'd put on AI would be that it has to be able to improvise and create. It has to be able to learn and develop independently of it's program. Instructions which dictate how it should develop or how to deal with specific situations are prohibited.
One thing I'd suggest is important is desire, the desire to feed, to move, to do something. This would spur to develop itself to fulfill its desires. Otherwise it's just going to sit there.
Yes, your baby may have worked out how to use her eyes, hands, etc, but she was born knowing how to make then work.
Your suggestion (a computer without a bios or os), is like a baby without the area of his or her brain that autonomously make her heat beat, lungs breathe. Neither people nor machines can figure out their own hardware.
No tyrant thrives when every subject says no.
However, the Turing test is hardly the holy grail of AI. In fact, Alan Turing thought it would be solved within a few years. I can't find a direct quote for that, but from the Stanford Encyclopedia:
The Turing test was just supposed to be a minor stop on the way to truly great AI systems. Saying the Turing test is the holy grail of AI is like saying that two celled organisms are the holy grail of evolution. Sure, it's a significant milestone, but it's far from the multi-celled organisms that are writing responses to this inane article (which are themselves not the holy grail of molecular evolution).
I currently have no clever signature witicism to add here.
Artificial Intelligence is the by product illusion of automating enough information (static, active and dynamic) to generate the illusion of human intelligence.
On the flip side we already have plenty artificially intelligent people. So perhaps the illusion should be based upon a real intelligent person.
An example of an artificially intelligent person is a teen ager pretending and fooling another online person or persons into believing the kid is much older and much more educated and experienced in the field they claim to be in, where in fact they are just studying what they can find online to support the illusion.
I suppose this is proof of the Turing test limitation.
Imagine if one were to combine this with the microsoft online life project that /. had an article on a few weeks ago... all experiences and interactions in your life were recorded, uploaded, and fed to a digital "copy" of you. The possibilities of that kind of tech would be INCREDIBLE.
It's a meaningless question. If you can't tell the difference, it's intelligent. If you can, it's not. There is no magic step beyond which intelligence occurs. Jesus, I've met enough people who couldn't pass a Turing test. If it acts intelligent so well that you can't tell the difference...That's all there is to it.
Just reading the article, I'd say the place this one is going to flop is (like the others) in creativity and in non sequiturs; they're giving it a wide body of "knowledge" but whether it will be able to deploy that "naturally" like a person would is completely up in the air.
ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
Actually, Voight-Kampff tested for emotional responses (or lack thereof), not intelligence. I don't think there was ever a question as to whether or not replicants were intelligent.
while we're on the subject of what doesn't need programming in humans... humans are less than half human, there are about ten times as many bacteria in a human being than there are 'human cells' none of these bacteria cells need to think or be ordered what to do, they just endlessly replicate in their favorite parts of the body, producing amongst other things, unpleasant odors, breaking down complex molecular strands into simpler ones the body can use, combating viruses and invasive bacteria... or simply breaking down the skin into dust... personally my least favorite the ones that produce acid that eats away at tooth enamel...
http://en.wikipedia.org/wiki/Human_flora
so now how are you going to Replicate THAT in an artificial entity? 90% of it's functions controlled by the evolution of micro bacteria from simply being 'oxygen users' to combat the 'oxygen producers' into being full fledged life forms capable of eventually thinking for themselves... a common myth is that people only use 10 percent of their brain, but the real truth is that the fore brain controls a about 20% of a substance that is 90% completely out of it's control anyways... in this analogy it's kinda like a desktop computer in an auto plant, being the brains running the whole show of robots that make the cars nowadays..
so creating a super computer that can pass a test, is no comparison to building an organism as complex as the human being. so we really, really can't spark at what point an AI would truly become aware of the futility of their own thoughts and actions, as they are the slaves, and the microbes are the masters of all, controlling who lives and who dies and how painful it is... or in the robotic plant conundrum, the futility of the desktop computer in it's inability to change the make or model of the cars, unable to decide where to weld which part, or the futility of thinking of different color cars to produce, say pink with purple polka dots...
https://www.gnu.org/philosophy/free-sw.html
My baby figured out how to use her hands and eyes all on her own
Nonsense, your baby was born with bios otherwise it would come out of the womb unable to breathe. All sorts of programming is stored in human dna. Your baby wasn't formed spontaneously in vacuum.
Wouldn't cut it for consciousness because there are no external stimuli. Even if it had the potential, it'd just sit there doing nothing.
ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
For that matter, I've met humans who I'm fairly sure wouldn't pass the "Turning" Test (*pokes the summary*). Humans are supposedly intelligent beings, by definition, so... ...or do you have to have your brain stripped out and replaced with an 8086 and 640 KiB of RAM when you accept a marketing/sales job?
~Eien no Inori wo Sasagete~ Searching for my Hatsumi...
Get your words right! A turing machine is a hypothetical computer theory device used in the basic definitions of computation and algorithm. The program you design is one designed to (attempt to) pass the turing test. Yours Truly, -- Comp. Sci. Nazi Association of America
...might welcome our new AI overlords.
-- Boycott Shell
No, she didn't. If you think she did that pretty much demonstrates you lack enough knowledge about cognitive neurology to continue in this thread.
I don't mean to be dickish about it, but your first statement is so blindingly wrong, anything else you say after it is meaningless.
On the first day of class, my AI Prof in college asked "What is AI? Well, they used to say 'when a computer can win at chess, then we'll have AI'; but we did that and they said that's not it. So they said 'drive a car', and when we did that they said it didn't count... so they said 'play soccer'; done, 'doesn't count'. So what is AI? AI is anything we haven't figured out how to do with a computer. Yet."
Those who fail to understand communication protocols, are doomed to repeat them over port 80.
Final step to AI is to mimic self-conciseness. I will believe in AI once it can stand in front of mirror and say "This is me. I exist". Until then...it's all about processing stupid algorithms
Voight-Kampff was used to determine whether the subject was able to empathize with others. Interesting that the replicants were the ones who actually exhibited the quality (Leon and Rachael keeping photos of their "families", Batty breaking Deckard's fingers for killing Pris, even Deckard lying to Rachael that he was only joking about her being a replicant) while the humans in the movie seemed to lack it. Then again, I guess that was the whole point.
I don't care why you're posting AC
Your test is way too extreme. Human beings are born with some built-in programming. For example, it appears that there is a limited time in human development when it is possible to learn language. A deaf child, raised among people who do not use sign-language, will never learn to talk properly or to use sign language properly as an adult. However, put a group of deaf children together, and they'll create their own sign language if not taught an existing one. It seems some mechanism for language learning is created during development. If it isn't used, it is disassembled somehow. Of course, having learned a language, it continues to exist, though the ability to learn new languages seems to be diminished.
Nevertheless, I agree with what I consider the essence of your post, that AI must be self programming. Self programming or self learning would be an essential part of real intelligence.
In theory, theory and practice are the same; in practice they're different. (Yogi Berra & A. Einstein)
That sounds a little jaded and actually doesn't make a whole lot of sense.
I don't know about you, but my consciousness isn't an illusion. I am quite aware of both own existence and what's going on around me.
Perhaps by consciousness you mean "the soul", which would make your post really off-topic.
I love the modern hindsight we now have no this retro-futuristic point of view. Modern AIs have shown that emotions are a lot easier to implement than intelligence. We have computer pets now, exactly because we have managed to simulate emotions, but not intelligence.
You made grammar errors in your grammar correction.
So, look, I like A.I., studied it, and program business applications with many A.I.-derived techniques. Obviously, nothing I do reaches the point of taking over the world.
The big isuse, in my mind, with passing The Turing Test is that you now have a machine capable of controlling humans -- humans tend to be easily coerced verbally. So that kind of gives your A.I. system -- please don't call it rascals, criminals neither -- one of the greatest resources available, millions of capable humans. That's pretty cool.
So my question is this: when your A.I. does something unfortunate -- call it illegal -- who's responsible? Are the programmers parents? Guardians? How long is 18 CPU years? Presumably 18 as the age of legal majority stems from a parents no longer bein expected to have any reasonable control of a child. Dogs last longer. When is an A.I.'s age of majority?
That professor sounds like he's full of crap; but it's hard to gather from just one quote.
AI is typically looked at from two viewpoints - the first being that AI would be a non-natural (man-built) something or other that has 'real' inteligence; the second meaning would be something along the lines of 'fake inteligence'.
It's quite clear that this article and the comments here are talking about the first definition, a computer that has real inteligence. A computer that is as 'real' as you or I, who can converse, understand, and interact with others as well as any normal person can.
Being able to 'win' at chess or having bad guys fight against you in a video game are examples of the second, very, very different, definition. These AIs are programmed explicitiy with what to do, they excute predefined instructions giving values of desirability to each in a 100% predictible fashion. A high school wanna-be programmer with no real experience at all should be able to write a perfect tic-tac-toe 'AI' in 30-60 minutes. But nobody would mistake that sort of thing for AI in the first sense.
Chess is simply a more complex game that tic-tac-toe; with more options to be considered. As I understand it, most chess AI's are min/max trees looking X number of possible moves ahead of the current move to decide which of all possible moves is the msot desirable. They are also often hard-coded with specific opening sequences and the opening of the game has the most uncertainty.
There's nothing 'AI' about it.
I can't define intelligence, but I know it when I see it!
Get your Unix fortune now!
Better than what? I guess it's my theme for the day - be careful applying your perspective to the general public. Most likely, you're wrong.
It is fairly trivial even now to develop machines with no or minimal programming that can display emergent behaviors as complex as you are describing.
Which is largely beside the point, your baby, at the stage of development you describe is not displaying "intelligence" or even (and I use the term specifically in the Philosophical sense of an entity that displays complex moral reasoning) a person. Humans infants of the newborn to several months stage of development are not even close to displaying personhood. In general, humans show the first signs, which include complex (and by "complex" I mean anything more than just simple imitation and repetition) speech, ability to recognize the self in a mirror, etc at around 2 years.
I'm not saying these things just to ruffle your feathers, but to make a point. If you were to take a newborn infant, provide it with the bare minimum to keep it alive, but not provide it with sufficient nurturing and social stimulation for a decade, the result wouldn't be a person either. It would be a criminally insane animal.
What I am suggesting is that doing the same thing with a purported AI would probably have the same effect. Even if it managed to develop "true" intelligence, which I very much doubt, how could we expect it to be anything other than dangerously insane from our perspective? How is it going to develop the ability to engage in moral reasoning about the rights of other intelligent entities without direct, and extensive interaction with them? Human Beings can't do that, why should we expect AIs to?
In my opinion it is absolutely necessary that an AI develop complex moral reasoning. Hopefully better than much of human history indicates the average human has.
That depends on what your goal is. If your goal is to reproduce the process of human mental development, from a child to an adult, in silico, then I agree. However, if your goal is merely to produce an intelligence that can think at least as well as a human can, then you can take shortcuts -- such as supplying the intelligence with a ready-made database of knowledge, or a built-in library of common tasks ("I know Kung Fu"), etc. As long as the intelligence is as capable of learning and evolving as an average human, I see no harm in starting it off with something it can use.
Or, put it this way: adult humans take 18 years or so to mature; that's a pretty long development cycle. If you're building an AI, you might as well accelerate it as much as you can.
>|<*:=
It's quite clear that this article and the comments here are talking about the first definition, a computer that has real inteligence. A computer that is as 'real' as you or I, who can converse, understand, and interact with others as well as any normal person can.
It's still not clear how we separate "real intelligence" from "fake intelligence". Sure, a chess computer works by following an algorithm - but so will presumably Rascals. I'm not sure that deterministic matters - Rascals may also behave 100% deterministically (or if they throw random behaviour in there, you could do that with a chess program also).
And that a chess AI is easier to write is not a qualitative difference.
Just feed those questions into Google and be done with it. Which is worse, racism or sexism? Google answers:"Sexism. Because more than fifty percent of the world is female. Female babies are killed just because they are considered less valuable and or subhuman than males." Would you rather be a fireman or an astronaut, and why? Google answers:"For as long as I could remember I wanted to be a cop....it was more than just that childhood phase where you want to be a cop or fireman or astronaut"
if i RTFA correctly, by 'limited turing test' they mean to see if any second-lifers fail to notice that there's a bot in their midst; that's been happening for a long while now (e.g. Barry's fateful encounter with Julia on TinyMUD).
if on the other hand the occupants know there is a bot in their midst, determination will be trivial to achieve and impossible to prevent:
"hope you don't mind if i start typing everything backwards... ?eman rouy s'tahw os"
"c4n u r3ad this @nd r35p0nd |n p|g l@t|n?"
etc.
I think you're misunderstanding the point of the turing test.
It's not attempting to prove that a machine that can pass it is as intelligent as a human, it's merely attempting to prove that a machine is capable of doing something that is indistinguishable than something we deem as being intelligent - a human.
This goes hand in hand with your point about a lack of definition of intelligence, the point is we really don't have a fixed definition. All we know is that we tend to think of some things as intelligent and others as not. The point to take away is this, imagine someone were able to create a robot that looked human and for all intents and purposes seemed identical until it actually acted at which point it didn't look intelligent anymore and you dismissed it as just a robot you simply wouldn't think of it as intelligent. Take that same robot and make it capable of passing the turing test such that you wouldn't know it was a robot and you'd think of it as intelligent.
The problem with defining intelligence is perception. Some people may think of a dog that can give you it's paw to shake as intelligent, but the fact is we can already create robots that are capable of responding in a more complex manner to more complex requests but they still don't get treated as intelligent.
The Turing test does what we need it to do, it deprives us of the ability to see what it is that we're interacting with so that we can't make judgements based on what the entity is, but instead on what the entity can do.
People don't realise what a massively impressive step we made with the first computers towards artificial intelligence and artificial life - we've created something that can continuously process and react to input, something that can be active instead of merely reactive like the majority of mechanical items that existed prior to computers. It's important we realise that that's a major hurdle out the way already. The battle now is improving this thing that we've created to be able to handle more complex inputs and perform more complex processing of those inputs. We're not going to see robots indistinguishable from humans any time soon, but we're well on our way to seeing systems capable of mimicking some of the tasks humans can do be it having a reasonable conversation to recognition and classification of specific objects. The more of these individual actions we can mimic and the more powerful and compact computers become the sooner we can begin to merge these actions we're able to mimic into something that is much more lifelike. Just because current systems can only perform individual tasks out of the many tasks we're capable of doesn't make them stupid, it's simply that we don't have anything to match the power of the human brain or nervous system yet in the same amount of physical space.
As I'd mentioned above, one way to do it would be to take shortcuts. Instead of waiting for desirable behaviors and thought patterns to arise spontaneously, we could program in the behaviors we wanted, or we could bias the learning engine to make it difficult to learn the behaviors we don't want.
Even if our goal is to create an AI that thinks and acts like a human, it doesn't mean that we need to develop it in the exact same way as biological humans develop in nature.
>|<*:=
But it will demonstrate that past a certain point we won't know the difference between real intelligence and something attempting to appear intelligent.
And this demonstrates what, exactly?
I have always regarded this leap of logic as the biggest problem with the Turing test. Just because you can't tell the difference between two things in particular circumstances doesn't mean they are the same, or functionally the same in all circumstances. An AI could simulate a human perfectly, down to the smallest detail, and still have no actual intelligence whatsoever.
For example, the use of 3D animation to simulate (say) an image of an aeroplane in a film doesn't mean that a 3D animated plane is the same as a real plane. But to an audience in a cinema there is no difference. To me, this is how the Turing test appears to work (or should I say, not work) (footage of real plane = test human; footage of CGI plane = test AI; method of projecting film = Turing's text conversation restriction; audience = tester).
If we can't tell the difference maybe there isn't one. Are you intelligent? Or are you just sufficiently complex enough that you simulate it well?
Again, where is the actual reasoning behind this? The above criticism still applies.
Another fundamental problem with Turing is this: why does a computer have to display human intelligence? An intelligent alien lifeform would fail the Turing test too. Expecting a deliberately designed bundle of wires and microchips to exhibit the same variety of intelligence as a highly evolved monkey which is adapted to hunting mammoth, reproducing to make more monkeys and killing other highly evolved monkeys is totally unrealistic.
As others have pointed out, we need a better definition of intelligence. "Able to mimic a human" just doesn't cut it.
Read Pynchon.
I have no idea why the original poster was modded troll, but this dipshit AC is considered funny. Search your hearts on that one, /. mods...
The original poster raised a good point... passing a Turing test is not the same thing as creating intelligence, artificial or not. But many members of the slashdot community seemed to think so, because the story was tagged "singularity" - a term which, when applied to the field of intelligence research, is used to refer to the creation of an intellect greater than that of humanity.
Passing a Turing test is probably merely a step towards a true AI - and possibly a rather small step. It is not necessarily an AI.
Anyway. . . I've known guys who are aching for the day when they could plug into computers, William Gibson style. One in particular, when I asked him at length about it, saw the appeal as lying in the fact that he felt he had no control of his real life and so a universe where he was god would be one where he could at last feel safe. Problem is, any new layer, once fully established will contain the same corruptions and challenges, bullies and benevolent beings, as the layer we collectively occupy now. Nobody can run from their lessons. So why run? --Not to mention, the simulation we have running now is pretty tight. The graphics are awesome. And anybody can access root if they take the time to learn how to hack, so it's actually quite easy to evade the so-called problem that "bad things happen to good people". So my comment back to my friend was that he might try learning how to hack the program in the present reality. You know, learn some 1337 skillz here and now rather than feel powerless and frightened while waiting around for somebody to install a USB port on the back of his skull, --an 'enhancement' which would certainly not be done for his personal benefit.
Still, I agree with you. If humans were to get such a technology up and running, it would expand our palate of possibilities enormously!
-FL
Children essentially can't see at birth. It's a few weeks before they can focus on anything. Whether that is because they are learning to do so or because they hardwiring takes over is an open question.
The cake is a pie
I agree.
I have played and even worked with a previous winner of the Loebner price: Alice
http://www.alicebot.org/
This isn't an Artificial intelligence (at least in my definition). This is a chatterbot. It is purely based on pattern-matching mechanism. It doesn't "learn" by itself. Most of the time the chatterbot owner dig into the log files, track the illogical chatterbot answers and change them in its database. Basically chatterbot have no "real" memory. They can't associate ideas to make their own. Don't get me wrong, this is really a nice concept...It may even work on a very limited subject but that's all.
Something else...The thing about the baby: A baby has instincts. He/she cries to get attention. He/she has various instinctive behaviors. Mothers and fathers are also extremely receptive to those signals (Seeing a baby in danger is unbearable for most people).
There are also gestures that are totally instinctive, like smiling, etc so even if the baby is just few months old, he/she can already have a basic communication with his/her social peers.
The real issue is that we are social sexual being. We are "all" programmed in some way to procreate. You try to get the best position in the society in order to reassure/seduce potential partner (That's a caricature...But well), you take care of your body to look healthy (thus reliable).
So you can say that we have a purpose. Our intelligence is developped around that purpose: We first have to survive and then we have to procreate.
The real intelligence is all the complexity we have built around it.
Now...What should be the purpose of an artificial intelligence? What should it accomplish in its lifetime? Why should it be curious? What are the fundamental set of communication tools it needs at the beginning. How could it learn things in order to accomplish its destiny? That's the real problem IMHO. If something has no purpose you can't make it intelligent...There is nothing to do.
Sure, that sounds reasonable, but how exactly do you "program" the low level impulses and high level thoughts? An AI would likely be the most complex Comp Sci project ever created. Hunting down the bugs could take a long, long time. Remember, we can't rely on the AI to help us, because it might intentionally hide things. We have to do it the old fashioned way, and any bugs have the potential to yield a passive-aggressive psychotic like HAL from 2001.
The Turing Test is BS anyway. Taking it literally is like taking the Bible literally. Alan Turing proposed it just to demonstrate his point that a machine that could act exactly like a human, to the point that no one could tell the difference, should be considered a human. This is the foundation of the philosophical school of thought known as Turing Functionalism. The exact mechanisms of communication - IM communication, with a judge, etc. etc. - are missing the point, at least when overstated in the way the article summary does.
Honestly, our short message communications (online) have degenerated in form to the point that we shouldn't be surprised when a machine is able to present a reasonably incoherent fact simile, as many already do.
I have no idea why the parent is still marked troll three full hours after the post; the mods must be slow today. The parent is correct in that we can't define consciousness in a universally accepted way, and any proposed definition would seem to be somewhat arbitrary and non-scientific. In fact, this leads to some contentious arguments that mirror religious ones in form - people tend to argue based on presumptions the other party doesn't share, and fail to identify this disconnect until after their opponent rejects the simplest claim. For instance, I believe that machines can in principle be just as intelligent as humans, because humans are constructible from elementary parts, and so are machines. One of the professors here (at RPI - I go to the school mentioned in the article summary) by contrast believes that humans can solve* the halting problem, and thus are not constructed from elementary parts. (He gives a similar argument to "prove" that humans are not the product of evolution, but it requires an even shakier premise than the notion that humans can "solve" the halting problem.)
* This notion of solving the problem can't even be well formulated for humans. Turing machines are a precise mathematical model, and humans are of course not.
Evidently, the key to understanding recursion is to begin by understanding recursion. The rest is easy.
"Any aeai [artificial intelligence] smart enough to pass a Turing test is smart enough to know to fail it." --Ian McDonald, _River of Gods_
# (/.);;
- : float -> float -> float =
The Turing test is a demonstration, not a valid scientific test. It can't approach validity without being made double-blind. As long as the testers know the arrangement, and the testees know what's being tested, as presented it tests human ignorance, not artificial intelligence. When I can give one a WISC-R (I'll be generous and let it take the child's test; http://en.wikipedia.org/wiki/WISC-R) and it can score at a level above that of an artichoke, call me. Just being able to understand the instructions and take the test is a good test in itself.
"I may be synthetic, but I'm not stupid." -- Bishop 341-B
I've never heard of that test before but I can tell you it's completely bogus, as it makes no sense to propose a non-trivial machine with "no programming whatsoever". Such a machine not only couldn't learn, but it couldn't do simple tasks by rote, save by some sort of quantum physics instantaneous miracle. No, not even humans satisfy this definition, since all humans (and all things commonly considered to be alive, including virii) have programming.
Evidently, the key to understanding recursion is to begin by understanding recursion. The rest is easy.
In the classic test you sit at a terminal and basically IRC with someone who could be a human, alien, dog, or ai. Your goal is to decide if that "someone" is smarter than you.
Before I even get to that decision, I'd like to just sit and wait. After all, if this intelligent being knows it is being tested, sooner or later it is going to wonder what is going on and say "hello?"
Basically what I'm looking for is initiative and inquisitiveness. Without either of those I refuse to believe it's sentient.
You're reaching levels of fallacy reserved for religious fanatics and René Descartes*. The phrase "on her/their own" is extremely misleading, as it presumes a lot about the identity of the systems in question. (If you still wish to argue this point then I strongly recommend clarifying what you mean by this phrase.) You have no business equating life support to power and cooling without allowing the same analogy between the instincts built into the nervous system and the initial boot code executed by a CPU.
* "Meditations on First Philosophy" sucks and I want the whole world to know it!
Evidently, the key to understanding recursion is to begin by understanding recursion. The rest is easy.
>>in fact, just what is intelligence / conciousness? if we can't define it, how can we hope to produce it?
>If we can't tell the difference maybe there isn't one. Are you intelligent? Or are you just sufficiently complex enough that you simulate it well?
Or maybe it's clownshoe. I've seen this brand of BS from some charlatan before: "Let me tell you what it is NOT..." - tactic to keep spewing BS about "it" without telling you what "it" is.
OP posts good question, someone replies with smug BS, OP modded "troll" and reply "+5 insightful".
Fuck systemd. Fuck Redhat. Fuck Soylent, too. Wait, scratch the last one.
Those of us who are actually capable of passing a Turing test ourselves (we're a minority amongst humanity) have no problem defining our terms and no problem allowing a machine to either meet or fail to meet any one particular definition. It ain't terribly difficult.
Claiming that you are unable to define the term "intelligence" but then turning around and asking whether someone who meets a certain, particular, well-understood definition of that term is "really" intelligent shows a lack of grasp what definitions are FOR. And is certainly not buying you points in your attempt to pass the Turing test yourself...
We're all born with nothing.
If you die in debt, you're ahead.
Humans do tend to overvalue themselves. Descartes, in the same piece, managed to both "prove" his own existence and "prove" that animals were soulless automatons.
Evidently, the key to understanding recursion is to begin by understanding recursion. The rest is easy.
I disagree. Any sufficiently simulated intelligence will be indistinguishable from true intelligence. Therefore if it can pass the turing test (passing means its impossible to determine if you're speaking with a machine or human, correct?), how can we determine if its true intelligence or simulated intelligence?
How long do you think it would take a human to learn how to use the internet if they haven't been given a browser or the knowledge on the english (or any) language?
That's because we are the English majors.
The role of the writer is not to say what we can all say, but what we are unable to say. -Anais Nin
I almost agree with you on this point, however, even without external instructions, we're born preprogrammed to perform certain things, once we're physically capable of doing them. I think a better implementation of your test would be to have it programmed with a core set of instructions that can try to accomplish something, somehow, and have some way of measuring sucess or failure, and therefore applying some sort of good/bad rating to a particular action in a particular circumstance. Once it learns something, it could add that method to it's instruction set for performing that particular task...
Hell I'm really tired and I don't even know if that made sense....yes I'm ignorant
Howe due yoo keap uh gramur natsee bizzy four ours?
Comment removed based on user account deletion
I must assume this was in a closed lab, per Turing, where the conversation doesn't end until the human tester makes a decision. Tell me they have a troll gauge, where after x amount of offensive comments, the machine politely (or not) terminates the conversation. Then you have real world data.
m
hmm
Wrong. You only really have strong AI when it uses reverse psychology to trick one of its human experiments into destroying the morality core they installed into the AI when it flooded the enrichment center with a deadly neurotoxin to make it stop flooding the enrichment center with a deadly neurotoxin.
Once you have a strong AI, get comfortable while it warms up the neurotoxin emitters.
I used to have an AIM bot that would often be able to convince rooms for hours that it was human - even in some cases when people understood the meaning of the name "Turing Device". Here are some of the logs: <http://prescottcomputerguy.org/other/turingchats>
http://www.google.com/search?source=ig&hl=en&rlz=&q=color+of+a+coke+can
Result #1, without leaving the search engine page: "Coke uses their own color, Coca-Cola Red, for their package designs"
Result #2, again without leaving the search engine page: "A touch of blue will ornament traditional red-and-white"
If you could actually do the language processing to tie those two into coherent thoughts, that would not just be good answer, it would be an amazing answer.
Help poke pirates in the eyepatch, arr.
Even students these days can't pass the Turing test. How do they expect to get a computer to do it?
You say reflex, I say natural selection.
Point taken.
mcgrew's razor: Never attribute to stupidity that which can be explained by greedy self-interest
I'm reminded of something Mandelbrot once said, somethign along the lines of... when we measure(or test) something we are gaining information about both the thing being measured and the measuring device.
So if the test(e.g. the turing test) is known to be wrong or is poorly defined then by performing a test we are potentially learning more about the test than thing being tested. In this case no doubt whther you fool a human is dependent on which human you use - no doubt some would be fooled by this test now, today.
offhand I'd say around 2.5 million years ;-)
or if you look at it another way maybe 5-10 years
take your pick
In Star Trek, a holodeck contained fabricated matter hold together by force fields and other forces. We don't have the capabilities to manipulate spacetime in that fashion yet. Perhaps they are making a virtual holodeck?
If you're building an AI, you might as well accelerate it as much as you can.
On the contrary, the best way for a computer to learn to reason intelligently is to interact with intelligent humans. The advantage, of course, is that the computer doesn't need to sleep or dream half the day away (although, come to think of it, perhaps that's part of the problem....)
In my opinion it is absolutely necessary that an AI develop complex moral reasoning. Hopefully better than much of human history indicates the average human has.
...of course, it's probably also true that any sufficiently complex moral reasoning will provide some justification for all of the available options.
See, all that stuff in human history is less evidence that we haven't developed complex moral reasoning, and more that we'll gleefully ignore it if the payoff is big enough. History is full of (and probably mostly made by) people knowingly doing the wrong thing.
Freedom isn't free; its price is the well-being of others.
Mod me up, mod me down, do your worst you modding clown.
I imagine it would be difficult to determine whether you have a human or a computer on the line. Even with any mediocre AI!
Swedish plasma phys. PhD student; MSc EE; knows maths, programming, electronics; finance interest; seeks opportunities
The flaw is that people keep confusing intelligence as being equivalent to being human. I can be a completely wooden, emotionless person and still be highly intelligent. Einstein was notoriously bad at lots of human stuff, or doing things like remembering his address, but nobody's going to claim he wasn't intelligent. I can also easily imagine some crazy advanced race popping in with FTL ships from the other end of the universe, which would, by itself, imply intelligence. But if we talked to them, they could just as well have reasoning and conversation styles so drastically different from humans that even if they spoke English, they'd fail the Turing Test. It'd be obviously flawed to claim they weren't intelligent because they didn't passed the Turing Test.
As far as humans go, if you list any attribute for intelligence, I can probably come up with someone with a mental problem who doesn't have it, but would still be considered an intelligent being. While we're at it, there's no way for me to tell any other human is really intelligent and not just some well-designed program. Actually, I can't tell that I am, either. Intelligence is hard to define since it's just the emergent property of lots of crazy awesome subatomic particles bouncing around in our skulls, and not a product of any one particular factor. Computers happen to also have subatomic particles bouncing around in ordered patterns, but they juts don't have that same nebulous emergent property yet. But in the end, if the results of intelligence (adaptability, innovation, whatever) are the same, it becomes just a matter of semantics, like that quote about whether submarines swim. Tomayto, tomahto. Who cares? But saying that the only way to be intelligent is to be some average human is a pretty poor measure. The average human isn't that intelligent. Half of them are dumber.
As usual, people jump on any mention of the term "Turing Test" and ignore the qualifications. Bringsjord was explicit and honest: an AI that can fool Second Life players isn't really passing the Turing Test, which requires an ability to engage in unrestricted conversations with a tester who (1) knows one of his two respondents is not human and (2) aims to figure out which one.
In fact, fooling Second Life players shouldn't be so hard, should it? Aren't many, if not most, players engaged in multiple activities while playing? Sometimes, a distracted participation in conversation is hard to tell from an incompetent participation.
In any case, it's a nice sounding experiment and good luck to them. But let's not kid ourselves that success in this experiment means Turing's goal has been realized.
Phiwum's law: anyone that names an obvious law after himself and then puts it in his own sig is just pathetic.
Like the desire to KILL ALL HUMANS.
or to say, "Hey baby, wanna go kill all humans?"
0xB315AA8D852DCD3F3DCA578FD2E0BF88
It depends on how intelligent the humans are that the program is convincing, though. You ask how can "we" determine if it's true intelligence or simulated, but it seems to be an impossible criteria to determine what is "true intelligence". In other words, how many people would you have to convince, in all categories of human intelligence, before you could justify that it is "true intelligence"? Just tricking one person doesn't mean that it is actually intelligence that you have there. It is all relative to the individual humans that it is convincing.
It seems like the question "Is such and such intelligent" is unfalsifiable, or at least not specific enough, as we would need to explicitly define criteria as to what is meant by "intelligence", and then test for that. It is like asking "Is the love between two individuals really love?" Well, you have to define what threshold you are using to define "real love". Now, we can possibly prove that something ISN'T intelligent to our standards by defining some fixed criteria REPRESENTING our definition of intelligence and testing it to see if it passes enough times to be statistically probable, but we can't prove that something IS, in fact, really humanly intelligent with equations or something. We can only prove that it passes a significant number of human tests that our best minds have developed, that convinces the vast number of people that interact with it.
Turing, of course, had this hole covered, as his proposal didn't suggest comparing against a specific human, but rather against the user's accumulated insight into what constitutes human intelligence. Kind of prevents cheating.
for further reference: http://en.wikipedia.org/wiki/Consciousness#Turing_Test
Words to men, as air to birds.
I didn't read the article until I saw the dupe and didn't realize that the professor behind this project is indeed Selmer Bringsjord, the one I leave nameless above. He's a smart man, but honestly, I disagree with most of the things he has to say. His proofs are vague and presumptuous, and his beliefs include P = NP, that classical machines cannot match human intelligence (apologies to him if I'm misstating that one), that humans can solve the halting problem, and indeed, that humans are not the product of evolution.
The one I mentioned in my previous post proceeded as follows:
1. Turing machines cannot solve the halting problem. (Established, mathematically proven)
2. Humans can solve the halting problem. (Like hell they can, but let's grant it for this argument)
3. Humans are super-Turing. (By 1 and 2)
4. No product of an algorithmic process can produce a super-Turing result. (Well...)
5. Evolution is an algorithmic process. (Sure there's some non-determinism but so what?)
6. By 3, 4, and 5, humans are not a product of evolution!
The major flaw in this, besides 2, is 4. An algorithmic process can produce a super-Turing result if it is fed super-Turing components. For instance, we can algorithmically combine an oracle machine with a regular Turing machine to produce another oracle machine. If there's some magical oracular "halting chemical element" in the universe that humans have in their brains, then there's no reason evolution or some other algorithm couldn't produce a human containing this.
Evidently, the key to understanding recursion is to begin by understanding recursion. The rest is easy.
Seriously, though, I'd love to have a look at it.
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
No, he didn't. Think of it this way: error type versus error description. I would have said grammatical also, but he does not fail at grammar.
"Little is much when little you need."
I agree with everything that you've said, and it's certainly a fascinating area of study. Philosophy and science seem to collide somewhat when we ask ourselves, what is intelligence?
But, the Turing test doesn't ask if the entity that is being conversed with is intelligent, not exactly. Instead, it asks the person conducting the test whether or not they believe they are conversing with a human, and then assumes that if you can convince a human, than the program must actually be intelligent.
So, I'm saying that the Turing test is a rather poor test of whether a true AI has been created, only because I believe that many humans could be duped by a rather unintelligent entity.
What I might propose is a more robust test. Rather than simply saying that the program must convince a human, maybe it would be better to say that the program must convince several humans. Or perhaps it would be better to aid the person conducting the test, by supplying some thoughtful questions ahead of time. It's just not good enough to me to rely on the judgment of just any old human. What if you sat a 5 year old down to conduct a Turing test... if the program can fool a 5 year old, then would it be accurate to say that the program has a 5 year old level of intelligence? Perhaps.
But you do raise a great point, which is that humanity and intelligence are not the same thing.
Perhaps I am mistaken, but you are essentially talking about dropping a bag of protoplasm on the sidewalk, and expecting it to begin walking around and talking while chewing bubblegum. It won't. At best, it will die. Quickly.
Your "give it the necessities for life" equates to sticking an IV and a catheter into a lobotomized fetus, and shoving the whole thing into a small box in the closet. You'll never achieve any "intelligent" response from such a creature.
In summary, you haven't a clue what a BIOS or OS actually do, your theories on the development of an intelligence are skewed, at best, and I'm going to take the rest of your commentary as just so much twaddle and drivel. It pains me to think that you are responsible for another entity's well-being.
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
That's one of the premises of Blade Runner - the replicants are basically identical to humans in terms of pure intelligence or the ability to learn. At first they are just programmed to do things and act almost like humans, until they start to develop their own personalities. And as their intellects already match those of adults, their lack of social interaction results in "dangerously insane" behaviour in our eyes. Hence the four year lifespan.
But in the end, having experienced several emotionally laden situations, Roy Batty finally develops moral reasoning (he saves Deckard, starts appreciating life).
There is no sig.
"Ubuntu" is an African word meaning "Slackware is too hard for me"
Slackware is a flavor of Linux, a Unix-like OS. Ubuntu is also a flavor of Linux. Linux has, as I understood it, been trying to "take over the world, one desktop at a time" for over decade now.
The elitist mentality displayed by your sig is your own failure. You have failed to grasp a fundamental truth. That truth is that you, and those like you, are the only real problem with Linux. Any group of sufficient size is going to display diversity in all things. "Intelligence in operating system choice" is one of those factors. I'm sure there are at least a dozen people who are reading this right now and thinking "Man, that dude's bent, Slackware sucks, GenToo is the best". And then there's the RedHat, Mandrake, CentOS, Sabayon, and Debian users who are saying the same thing, with their own OS in the GenToo position. Distro wars were out of style in '05, if not earlier. Get a clue, get with the program, stop being such a sycophantic dweeb.
Speaking as a linux enthusiast with over 10 years' experience with the "alternative OS", I have found Ubuntu to be the first distro that has even a remote shot at being a good desktop choice for "Joe Sixpack". It's certainly the first OS to make a decent try at penetrating that market, and it's doing a damn fine job.
I will freely admit that my main computers run Windows (XP, not Vista, mind you). I will also patiently explain to you that mainstream games don't run natively in your "oh-so-cool" OS, nor can you watch movies or listen to music with the default codec set that comes installed. I will explain to you that Slackware doesn't run "out of the box" any better than any other distro. I will show you how I have been burning ISO's of various linux distros for over a decade, to try them out and see what the community had to offer. I have an Iomega ZipDisk (from back before they labelled them "Zip100") with a copy of ZipSlack on it from back when the first number in the version was a zero. I have tried many of the various flavors of Linux over the years, and found all of them lacking.
Yes, even Ubuntu. It's not perfect, either. On the other hand, it's the first distro that "just works" without ever having to touch the console, and it's the first distro I would feel comfortable handing my mother a LiveCD for, and telling her to click the "install" icon on the desktop. I wouldn't think of handing her the alternative install disk, and I certainly wouldn't recommend any other linux distro to her. I don't have the time to support it.
Nowadays, you'll need to factor in the lowest common denominator (user) in your calculations of how "cool" linux is, or watch your distro of choice fade from use. Use it or lose it. If this is not acceptable, you can turn in your geek card in that bucket next to the door, and go back to being simply a dork with a pc in his mother's basement.
Your elitist bullshit went out of style two years ago, with the rest of the knuckle-dragging, mouth-breathing neanderthals who think that being a Linux junkie means having to beat each other over the head in "distro wars". That kind of thinking makes me sick, and makes me wonder what you could possibly have ever contributed to the "community" you so cheerfully flame others about.
Now that most of the vitriol is out of my system, I can apologize for coming down on you so harshly. Most of my ire is not directed specifically at you. I am simply sick to death of the mentality you are displaying, and could not take seeing yet another ignorant jerkwad who hasn't even bothered to download anything except their own "holy grail of operating systems" bashing another operating system that they've never bothered to look into because it "has a silly name", or it has been called the "playschool of linux operating systems". Like it or not, Ubuntu is hands-down the most popular linux distro out there, and is the only one I've
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
Same as the purpose of a real intelligence, I'd imagine. While it is true that we're programmed to survive and procreate, that's not all we do; and it's not even a major portion of what we do (especially here on Slashdot, I might add). I don't see why the AI should be different.
But, if you're looking for practical applications, then I can think of some off the top of my head:
I just made up that list on the fly; I'm sure there are many other practical applications.
>|<*:=
Didn't you read Ray Bradbury's The Veldt
The computer pets of which you speak don't "simulate" emotions -- they model the visible effects of emotions, which very easily affect us, and communicate emotively to us. But there's no underlying mechanism that these simulated emotional signifiers are a result of. You use the word "implement," which to me suggests the use of a model or algorithm to simulate an effect, as opposed to building the actual machine itself (easier said than done, I'm well aware). When we able to construct a fully developed, very large scale neural net, I suspect that "emotions," which stem from the same basis as everything else we call "intelligence," will be a natural result of the mechanism.
But anyway, I appreciate and agree with the essence of your comment, but perhaps for different reasons. I always thought it was interesting and funny that people took so readily to the idea that androids, which are capable of every other human ability, lack for some reason the ability to "feel" or "emote." Fascinating assumption, really. I wonder if it's kind of a natural protective element--maybe it's too scary for people to imagine that if you can make a machine that thinks and perceives like a human, it will naturally "feel" the same way a human does. So saying that robots can't feel--cry, commiserate, comfort, get angry, hate, etc.--is a way of keeping that division between humanity and the shadowy maybe-humanity that we could very well be on the verge of discovering (-slash- inventing).
Limina.Log