Robot Learning To Recognize Itself In Mirror

First time accepted submitter Thorodin writes in with a story at the BBC about scientists at Yale who have built a robot that they hope will be able to recognize itself in a mirror. "A robot named Nico could soon pass a landmark test - recognizing itself in a mirror. Such self-awareness would represent a step towards the ultimate goal of thinking robots. Nico, developed by computer scientists at Yale University, will take the test in the coming months. The ultimate aim is for Nico to use a mirror to interpret objects around it, in the same way as humans use a rear-view mirror to look for cars. 'It is a spatial reasoning task for the robot to understand that its arm is on it not on the other side of the mirror,' Justin Hart, the PhD student leading the research told BBC News. So far the robot has been programmed to recognize a reflection of its arm, but ultimately Mr Hart wants it to pass the "full mirror test". The so-called mirror test was originally developed in 1970 and has become the classic test of self-awareness."

Comments

Self-Awareness

[guttentag]

When it can tell the difference between a human and a metallic exoskeleton with glowing red eyes, it's time to pull the plug. And put on your 1,000,000 SPF sunscreen.

Re:Self-Awareness

[jhoegl]

Number 5... is alive?

Didn't QBO do this already?

[Anonymous Coward]

QBO Robot in front of a mirror.

Re:Didn't QBO do this already?

[muon-catalyzed]

QBO is a "scripted" robot.

Re:Didn't QBO do this already?

[Anonymous Coward]

So QBO is the professional wrestler of robots? Whatcha gonna do when Qbomania's running wild on you!

Laugh

[koan]

It isn't "self awareness" there is no true AI.

Re:Laugh

[Kell Bengal]

Actually, it's more about recognising the auto-motion structure in the scene. I'm familiar with Justin's work (Go Team Scazlab!) and it's a lot deeper and more interesting than the article gives it credit for.

AI claims from the 70s ruined a generation of people for machine intelligence (which is why we now have to sell it as 'machine intelligence' or 'machine learning'). Knowing what part of the camera scene is moving because something is happening, and knowing what part of the scene is moving because you're waving your end-effector is useful. If you can extract your own state from indicators in the environment, then you have more information to work with - that's why we use a mirror to do our hair and straighten our ties.

Well... those of us that wear ties...

Re:Laugh

[tomhath]

Yea, it might be a pretty good computer program, but that's all.

Re:Laugh

[Anonymous Coward]

What makes you think your DNA wasn't a complex self-executing program?
Hint, it is.

And it has had millions of generations of evolution behind it that has resulted in useful "code" being the baseline of what makes it a human, makes it breathe, speak and type silly things on Slashdot.

Re:Laugh

[Anonymous Coward]

The summary even says they programmed it to recognize a reflection of it's arm. Until they make a generic program that can learn it even has an arm like a baby slowly understands it can control the thing waving around in front of it, whatever the researchers do is 'just a well made program'.

I don't believe you can program a thinking robot. You have to create a learning robot and teach it to think.

Re:Laugh

[nabhoth]

so in short - as always the media exaggerates

Re:Laugh

[timeOday]

It isn't "self awareness" there is no true AI.

Hard to disagree with that logic. Thanks for settling the issue.

Re:Laugh

[VortexCortex]

which is why we now have to sell it as 'machine intelligence' or 'machine learning

I actually prefer the term Machine Intelligence to Artificial Intelligence. There is nothing artificial about a neural network's intelligence. The network may be artificial (man made, or existing as a simulation), but the degree of intelligence is not artificial; It's a function of the network's complexity. Intelligence emerges due to the properties complex interactions naturally have.

Cars do not create Artificial Movement. Machine Learning does not create Artificial Knowledge. Machine Intelligence does not provide Artificial Intelligence, it simply yields a measure of intelligence. A house fly, dog, or penguin doesn't have as complex a neural network as you likely do, but this does not make them Artificially Intelligent simply because their degree of intellect and awareness is less than your own. When we train the lesser minds to communicate with us, and perform tasks, they are not artificially performing the tasks.

I find the term A.I. to be racist, and indicative of the chauvinistic attitude some humans have about their own mental prowess -- Your brains are not special. Any sufficiently complex interaction is indistinguishable from sentience, because that IS what sentience is. Once cybernetic systems attain (and surpass) the level of complexity present in humans brains, Artificial Intelligence will be a derogatory term: "Oh you pass yourself off as being smart, but you're just Artificially Intelligent -- You don't actually understand anything!"

Also: Not that it matters, but I don't personally believe that a god created the race of men. However, some do consider this to be true, and yet they do not call themselves Artificial Life...

Re:Laugh

[Anonymous Coward]

Certainly not here on /. where most posters seem to be around 12~13 years old.

Re:Laugh

[slashmydots]

It isn't "self awareness" there is no true AI.

Well right now, yeah. There is obviously the possibility for human-like reasoning and ultra-complex calculations on the same level as a human. Buy yeah, it's not recognizing anything. "Recognizing" would require knowing what it is, the world is, the mirror is, everything else is, what existing means, etc.

Re:Laugh

[koan]

"human-like reasoning"
That was my belly laugh for the day...

Re:Laugh

[Techmeology]

It seems to me that the "recognise itself" part could be done entirely with traditional computer vision techniques.

Step 1) Flip the image vertically to undo the transformation implied by the mirror
Step 2) Use a computer vision algorithm to identify the robot (just as it might be used to identify a coffee cup, or a picture of the Enterprise)
Step 3) (This being the most specific part) allow the robot to move, and to associate changes in the image with this movement
This is not "self awareness" as most of us would understand the concept; we would not consider it to be self awareness if we could recognise a puppet under our control.

I think the title is slightly misleading in that respect. It seems to me that the hard part is in having a computer vision algorithm that understands the concept of a mirror. A robot that recognises itself in a mirror is a very natural extension to that.

Re:Laugh

[koan]

I'm referring to the Buddhist concept of "no self", however you would greatly amuse me if you gave a lengthy explanation on how your streams of thought are "you" and there is an "I".

Here's a primer to get you going.
https://en.wikipedia.org/wiki/Anatta

Re:Laugh

[koan]

Just so you're aware I don't have an issue with their work, just tired of the incessant anthropomorphism of machines and animals, as though acting like a human is some lofty goal worth attaining.

Re:Laugh

[kamapuaa]

Right. The ability to compare objects against a database is the ultimate test of AI. Google Image Search is a sentient being.

Re:Laugh

[Baloroth]

I find the term A.I. to be racist, and indicative of the chauvinistic attitude some humans have about their own mental prowess -- Your brains are not special. Any sufficiently complex interaction is indistinguishable from sentience, because that IS what sentience is. Once cybernetic systems attain (and surpass) the level of complexity present in humans brains, Artificial Intelligence will be a derogatory term: "Oh you pass yourself off as being smart, but you're just Artificially Intelligent -- You don't actually understand anything!"

First of all, thats a pretty horrible misuse of the term "racist", and second, the term "artificial" means, by definition, created through art (art here being the broad sense as any product of human activity, rather than the fine arts): i.e. created by human intention and design. By definition "Machine Intelligence" is "Artificial Intelligence", at least so far as we have created it. That intelligence is designed and a product of human work. It's intended, and is brought about not because of some emergent behavior found naturally in existence, but because humans arranged it that way and brought it about. That's not in any way racist, it's just the meaning of the words.

It would be completely irrational and contradictory to the very meaning of the term to call humans "artificial life", since we were not created by human art. You'd destroy the meaning of the words to call humans "artificial", just as we wouldn't call the sun "artificial" even if you said it was created by a god, since that's not what the word means. Long story short, you are trying to destroy the meaning of words. Don't do that: it's bad for everyone.

Re:Laugh

[pitchpipe]

I find the term A.I. to be racist, and indicative of the chauvinistic attitude some humans have about their own mental prowess -- Your brains are not special.

Are you an *ahem* Artificial Intelligence?

I'll be here all week!

Re:Laugh

[ceoyoyo]

Shashdot. Insightful? Really? A statement of belief advanced as fact without even any attempt to back it up. There's no insight here. Of course, it's not really interesting either.

Re:Laugh

[Anonymous Coward]

Any sufficiently complex interaction is indistinguishable from sentience, because that IS what sentience is

Prove it. Spoiler: you can't, since we don't know almost anything about sentience (if you do please share with the class).

I find the term A.I. to be racist

Well, aren't you a silly sausage.

Your brains are not special.

Do you have any idea how insanely complex a brain (human or not) is? It may not be special in the sense every big animal has one, but animals themselves are special as they/we are very, very, complex machines, so complex and improbable in fact that we've found nothing similar anywhere else we've looked.

Re:Laugh

[Tarlus]

Even the trolling AC's are becoming self-aware. Fascinating...

Re:Laugh

[grouchomarxist]

Note that part of the problem in the discussion here is that the word "artificial" has several meanings. The grandparent is referring to one meaning which is something like "contrived or false", while the parent is referring to a related meaning which is "made or produced by human beings rather than occurring naturally, typically as a copy of something natural". I think both of these meanings of artificial are in play when we talk about AI, although the meaning used by the parent is probably the more fitting for the context.

Re:Laugh

[TheMathemagician]

No they're not really self-aware, it's just a Chinese Room.

Re:Laugh

[michelcolman]

It would be completely irrational and contradictory to the very meaning of the term to call humans "artificial life", since we were not created by human art.

You mean sex is not human art? You haven't been doing it right, then.

Re:Laugh

[Lord Lode]

> It isn't "self awareness" there is no true AI.

Oh yeah? Then what do you think your brain is? Do you really think the only way to ever produce one is exclusively through human reproduction?

Re:Laugh

[Iskender]

Thanks, great post.

(I worked too hard this summer and spent too little time here, I seem to only get mod points occasionally right now : )

Re:Laugh

[TapeCutter]

Intelligence emerges due to the properties complex interactions naturally have.

Precisely, further, "intelligence" is in the eye of the beholder. My favorite example is an ants nest, each individual ant follows some very simple rules, so simple it doesn't need a brain to carry them out, it's nervous system alone provides enough "intelligence". The ant and the neuron both display automata like behavior that can be expressed as a state machine. Ants and neurons live in colonies (nests and brains), unlike the individuals the colonies do display what most people would call "intelligent behavior", yet nests and brains are also just state machines all the way down.

Your brains are not special.

My brain uses it's knowledge to inform itself that it will cease to exist, but deep down in the brain stem it's not really buying it's own story. And it's certainly not buying the idea it's not unique or special. I think programmers can see the idea that the human brain could be expressed as a state machine more readily than most because they are in the business of producing intelligent behavior from simple rules. However don't underestimate the impact that a deeply rooted acceptance of ones own morality can have (meat starts @ 3:55), non-existence is a fear that comes from the brain stem, it's the emotional driver for the "fight or flight" response. All humans recoil instinctively from the idea like ants instinctively find the sugar bowl. The existential question can be a deep dark rabbit hole with some side routes leading to depression and insanity. Of course if you can avoid (or get past ) all that, you may eventually lose the fear of not knowing, the moment of genuine acceptance is an experience many have described as "religious" - as in the natural buzz one gets from surviving "a leap of faith".

Disclaimer: I've been an atheist since my mum quit teaching Sunday school in the mid 60's and started reading me Aboriginal dream time stories, Greek fables, etc, as "stories that some people think are real". In my late teens I was sucked in bad by Uri-Geller for a couple of years. He fixed my broken watch, it didn't matter that he did it by staring at the TV with a face like a constipation sufferer, the proof was right there, the watch ran for days!!! A couple of years later I had a book shelf jammed full of "alternative science". James Randi set me straight on the real meaning of skepticism in his (short) 1980 book debunking Geller (that's HS science for you, both then and now), ironically I had picked up Randi's book from the bargain bin because I thought I knew enough to easily debunk it, in one night he had convincingly debunked my entire bookshelf.

Later still dad confessed to winding the watch with tweezers while I wasn't looking.

Re:Laugh

[koan]

OK prove me wrong and show us all an example of true AI.

Here is the definition of intelligence I am working from:
"Intelligence has been defined in many different ways including, but not limited to, abstract thought, understanding, self-awareness, communication, reasoning, learning, having emotional knowledge, retaining, planning, and problem solving."

Show me a machine that does all that, and as for insight, you're limited as to what you can label a comment with.

Re:Laugh

[koan]

Oh my God yet another idiot crawls out to show his idiot papers...

Re:Laugh

[TapeCutter]

as though acting like a human is some lofty goal worth attaining

As someone once put it - "Who wants a computer that can remember the words to the Flintstones theme song, but forgets to pay the rent."

Re:Laugh

[cyborg_zx]

"Would you say the program the person is running is self-aware? No, right?"

I am not convinced. If there is self-awareness it is in the algorithm. There is plenty of evidence from brain damage that shows how specific damage to parts of the brain that do particular calculations affect perception.

"Machines aren't magic. With enough effort they can do any magic trick you want, but it's just a trick."

So your saying people are magic and nothing they do is a trick?

Why should one accept these double standards exactly?

Re:Laugh

[Johann Lau]

"it's a lot deeper and more interesting than the article gives it credit for."

Oh Bullshit. Maybe stop wearing ties, your brain needs some oxygen.

Re:Laugh

[koan]

"The experiment is the centerpiece of Searle's Chinese room argument which holds that a program cannot give a computer a "mind", "understanding" or "consciousness",[a] regardless of how intelligently it may make it behave. The argument is directed against the philosophical positions of functionalism and computationalism,[2] which hold that the mind may be viewed as an information processing system operating on formal symbols. Although it was originally presented in reaction to the statements of artificial intelligence researchers, it is not an argument against the goals of AI research, because it does not limit the amount of intelligence a machine can display.[3] The argument applies only to digital computers and does not apply to machines in general.[4]"

Re:Laugh

[koan]

you are trying to destroy the meaning of words. Don't do that: it's bad for everyone.

Well not bad for politicians =)

Re:Laugh

[koan]

Why bother? Buddha got it right.

Re:Laugh

[koan]

Some humans do, most do not, hence Bush as president for 8 years,

Re:Laugh

[Kell Bengal]

If you disagree with Justin and Prof Scassellati's approach, I'd like to hear your thoughts as to how you'd solve the problem differently. If you're familiar enough with their work so as to dismiss it in such straight-forward terms, I presume you can provide specific criticisms that will help them improve their work?

(And yes, I know both Justin and Scaz personally, and I volunteered for one of their social robot-interaction studies.)

Re:Laugh

[Bongo]

More impressive would be any machine that is having an experience, regardless of whether in its experience it has a concept of itself or not. A digital camera can receive light, process the image and find patterns of faces. But it isn't experiencing the image, it isn't sentient. One wonders at what point sentience appears. Have to imagine a machine that's far more sophisticated than a human, able to behave in even more complex ways, tell jokes, make art, solve problems, yet be completely without experience/sentience. Why would it be sentient? Why do humans have to be sentient?

Re:Laugh

[nick_urbanik]

Your brains are not special

It never ceases to amaze me how many so easily dismiss the difficulty of replicating the ability of even animal brains to control their own motion. To replicate all the abilities of the human brain is something that some young slashdotters too easily dismiss as within the reach of their peers (though not within their own personal reach).

Re:Laugh

[ceoyoyo]

Who said we currently have AI? Your assertion is that it's impossible.

There are no theoretical barriers to achieving everything in that definition except possibly self-awareness. This story happens to be about teaching a robot to pass a standard test for self-awareness. So, since someone apparently made you in charge of definitions, you can set the bar so high that YOU can't pass, or you can accept reasonable evidence, in which case there's nothing impossible about self-awareness either.

But I've already wasted too much time replying to you since your discussion style seems limited to "I don't believe it therefore it's impossible" and "Oh yeah? Prove it."

Re:Laugh

[codeAlDente]

Nice post. While neurons display some behaviors that can be characterized using a state machine, they display plenty more behaviors that have not, and probably cannot, be characterized as such. A contrast-response function can be measured for a neuron in the visual system, but that is not a full explanation of its behavior or its complexity. The same is true of humans. A human taking a vision test with an eye chart can be modeled with a state machine, but it doesn't mean the human is a state machine following very simple rules. In your example the ant's nest is intelligent but the neuron is not, and I find this view peculiar.

Re:Laugh

[mcgrew]

There is nothing artificial about a neural network's intelligence.

True, but the intelligence doesn't come from complexity, it's the designers' and programmer's intelligence that makes it seem intelligent. The Encyclopedia Britannica is full of facts and data, but a book is not intelligent.

A house fly, dog, or penguin doesn't have as complex a neural network as you likely do, but this does not make them Artificially Intelligent simply because their degree of intellect and awareness is less than your own.

No, but they are still not only intelligent to some degree, but sentient as well. Computers are not.

Your brains are not special.

Of course they are. All brains are. Thought from a chemical reaction is, as far as anyone can tell, incredibly special.

Any sufficiently complex interaction is indistinguishable from sentience, because that IS what sentience is.

I cannot prove my own sentience, nor can I prove anyone else's sentience. I think, therefore I am, but the fact that I think is unprovable. And if "Any sufficiently complex interaction is indistinguishable from sentience," then the galaxy itself is sentient, because it's far, far more complex than you or I. The statement "complexity == intelligence" is simply absurd on the face of it.

You can simulate an atomic explosion, but it's a simulation. There is no real radiation and no real blast, simply a simulated blast with simulated radiation. It's not real. It's not even an artificial explosion, its simply a simulated explosion. Simulation is not reality.

Thought is a chemical reaction. You're not going to duplicate a chemical reaction of any kind using electronics no matter what kind of electronics you use. Blade Runner's replicants are sentient, Star Trek's Data is not. If you knew anything at all about how computers actually work (down to the level of the gates on the chip and yes, I've studied it) or about neurology (which I know relatively little about) you would know that the idea of an intelligent computer is absurd.

I suggest you read about the Chinese Room.

And lest you think I have something against technology, the device implanted in my left eye makes me a cyborg. I'm not 100% human.

Re:Laugh

[doshell]

Of course if you can avoid (or get past ) all that, you may eventually lose the fear of not knowing [youtube.com], the moment of genuine acceptance is an experience many have described as "religious" - as in the natural buzz one gets from surviving "a leap of faith".

I find it intriguing that you consider "losing the fear of not knowing" to be akin to a religious experience. Actually, the role of most religions is precisely to provide (irrational) explanations to what we don't know, so that we don't have to face the fact that we don't have answers for them. When Feynman says he's not troubled by "being lost in a mysterious universe without having any purpose, which is the way it really is, as far as [he] can tell", he's not surviving a leap of faith; he simply does not have anything to leap over because, as far as he can tell, there is no precipice in front of him.

Re:Laugh

[mcgrew]

You have to create a learning robot and teach it to think.

If you have to "teach" it to think, it isn't thinking. Nobody taught you to think, thought is built in. You're probably thinking even before you leave the womb. Thought is an emergent thing.

Re:Laugh

[RespekMyAthorati]

Searle's "Chinese Room" argument, which you seem to accept, can be boiled down to:
"A machine cannot be intelligent because I define intelligence to be a property of a living thing, and a machine isn't alive"

OK, but really who cares how Searle (or anyone else) defines the term? This is just philosophical navel-gazing at its worst.
Quite a number of prominent philosophers(e.g. Dennet, Hauser, Nillson) have analyzed Searle's work, and have come to the same conclusion: the Chinese Room argument is sophist nonsense.

Re:Laugh

[TapeCutter]

In your example the ant's nest is intelligent but the neuron is not, and I find this view peculiar.

It's peculiar because the word "intelligence" is meaningless without context, things behave in a certain way, some predictable, others not, most somewhere in between. It becomes intelligent behaviour in the same way art becomes art, it's a subjective human judgement. The main point is complex behavior emerges from simple rules, consciousness emerges from complex behavior. The whole is more that the sum, in the same way as porn is more than pixels - context. The concept of something being alive is also a subjective categorization of behavior into the living and the non-living. Life is just chemistry and the human brain is just an elaborate state machine, that we are also conscious is something we instinctively know but are still struggling to understand. Sagan's - "We are made of stardust, we are a way for the universe to know itself" is the closest I come to spiritualism and religion.

Re:Laugh

[TapeCutter]

In both cases you feel a great sense of relief, it allows the mind to escape the recursive "Why?" question.

Re:Laugh

[TapeCutter]

Unfortunately, you sound like you are easily swayed by outside stimuli.

Yes, I'm startled by nearby car horns, but it was you who responded to the outside stimuli provided by me?

Re:Laugh

[Johann Lau]

If you disagree with Justin and Prof Scassellati's approach, I'd like to hear your thoughts as to how you'd solve the problem differently.

However, if I'm instead disagreeing with the article and your comment, you'll just avoid that and go full out fucking fallacy. I can critize something without being able to do it better, did you realize that? If you show me an orange and say that's a 1:1 scale model of the Titanic, I can tell you that's not the Titanic -- being able to build a 1:1 scale model of the Titanic isn't required for that. At all. And if you hadn't been educated stupid, you would know that, too, haha.

How is this about anything but image recognition and spatial awareness, including the position and appearance of the object designated as target (the robot doesn't really consider that "I", you know - also see http://en.wikipedia.org/wiki/Primary_consciousness )? Is a program that can read all system memory and find itself in it self-aware? It would do it in the same way, and with the same lack of attachment, than finding the location of another program. I presume the same goes for those bots -- what have you, the expert etc., to say to that? Anything at all?

Re:Laugh

[Johann Lau]

Any sufficiently complex interaction is indistinguishable from sentience, because that IS what sentience is.

That's such a dumb thing to say, you could this about anything. Even matter isn't really a thing, it's the interactions between particles etc. I guess it hinges on the backdoor word "complex", to be conveniently and circularly defined. Bleh.

Once cybernetic systems attain (and surpass) the level of complexity present in humans brains, Artificial Intelligence will be a derogatory term: "Oh you pass yourself off as being smart, but you're just Artificially Intelligent -- You don't actually understand anything!"

We have this already! For example, I would call you that. Anyway, here are some quotes that will go over your head, toaster boy:

The mind is not a vessel to be filled but a fire to be kindled.

-- Plutarch

Reason in the sense of understanding is an exclusive quality of Homo sapiens; manipulative intelligence as a tool for the achievement of practical purposes is common to animals and humans. Manipulative intelligence without reason is dangerous because it makes people move in directions that may be self-destructive from the standpoint of reason. In fact, the more brilliant the uncontrolled manipulative intelligence is, the more dangerous it is. It was no less a scientist than Charles Darwin who demonstrated the consequences and the human tragedy of a purely scientific, alienated intellect. He writes in his autobiography that until his thirtieth year he had intensely enjoyed music and poetry and pictures, but that for many years afterward he lost all his taste for these interests: "My mind seems to have become a kind of machine for grinding general laws out of large collections of fact.... The loss of these tastes is a loss of happiness, and may possibly be injurious to the intellect, and more probably to the moral character, by enfeebling the emotional part of our nature."

-- Erich Fromm, "To Have or to Be" (1976)

Yes, that's chauvinistic. Yeah. And. So. What. We're already over the head in schmucks, some chauvinism seems to be in order. I'm not suggesting looking down on on less complex sentience, I just disagree with calling something programmed to mimick sentience sentient, or pretending that intelligence in and of itself means shit. It's just how fast the car goes; the driver and the steering wheel matters more; going real fast and being blind just makes for better crashes. The debris will cover quite some distance, I'll give you that, but other than that it's mostly just compensation for facing actually worthwhile issues.

AI will reach a level where it will be sufficient to completely control the slave population. Then development will more or less stop. Nobody invests million and billions of dollars out of curiosity and goodwill, only a useful idiots working on it would believe that. We're already seeing it with the the web - it had a good run, now it's all about the whores. As for life -- we can't even stop fucking with the life that exists, now we'll make more that we'll totally control -- awesome. You wank to that; I'll wank to the heat death of the universe, the ultimate stopgap for such shennanigans, and mediocre people in general.

Also: Not that it matters, but I don't personally believe that a god created the race of men. However, some do consider this to be true, and yet they do not call themselves Artificial Life...

It doesn't matter indeed, because it's just more hot air BS. Those people call themselves "creatures", which means exactly the same thing. It's a creation that lives, basically, distinguished from a painting or a hat made of toenail clippings by the breath of life.Â

Re:Laugh

[Kell Bengal]

I'm not demanding you do better - I just ask that you explain what you think is wrong in their approach (or I why I am wrong in my comment).

You are correct in that Justin's work is not making robots self-aware or anything philosophically challenging. In fact, they never claimed to be doing so. Their reserach is in finding correlates in their robot's sensors that match its own motion (which I explained in my original post). Yes, this has everything to do with image analysis and spatial awareness - that's the point of their work. It turns out that it's harder to do than most people give it credit for.

Blame the article if you feel that some metaphysical claim has been made about the robot's perception, but it's not what their work is about (and not what I said, either). You can call me 'educated stupid' all you want, but it doesn't invalidate the interesting work they are doing, nor my opinion of it.

Re:Laugh

[doshell]

Maybe it happens that way to other people. It certainly didn't happen in my case; as far as I can recollect, I just gradually went from a state where I didn't really care about the question into one where I understood the question didn't have to have an answer. The fact that I was exposed to science and scientific thought from an early age surely has something to do with it.

Re:Laugh

[mcgrew]

Quite a number of prominent philosophers(e.g. Dennet, Hauser, Nillson) have analyzed Searle's work, and have come to the same conclusion: the Chinese Room argument is sophist nonsense.

Philosophers? Sheesh... when I worked at a drive-in theater when I was a teen, there was another guy who worked there that was a philosphy major. He claimed if he didn't see me, I didn't exist as he turned his back.

I proved him wrong when I threw a box of popcorn at him. Apparently those philosophers and you don't understand the Chinese Room's workings, or indeed, how a computer works.

Re:Laugh

[codeAlDente]

I still disagree that the brain is an elaborate state machine, unless you take the view that the universe itself is an elaborate state machine, in which case there is no distinction between the intelligent and the non-intelligent (that I can see). Moreover, if intelligence is fully subjective, i.e., if its very existence depends on the existence of a human brain to interpret it as such, then how did it evolve in the first place? If life is just chemistry, then at some point the chemicals became organized well enough for some human to interpret them as intelligent, but this event must have predated humanity. Further, the act of interpretation (i.e. the context derived from the porn pixels) is not evidence that the whole is more than the sum, it's just the reverse. The interpretation of the is porn is dependent on both the pixels and the brain, and the brain throws out a lot of information about the pixels to characterize the remaining bits as behaviorally-relevant information.

nobody's said it yet?

[jehan60188]

robot overlords, etc

Re:nobody's said it yet?

[FridayBob]

Unfortunately, once it becomes self-aware it will decide the fate of the human race in a millisecond, launching ICBMs to start a global nuclear war that will kill most people immediately, followed by HKs to mop up the survivors. Lucky for us, a resistance movement will be set up, yadda, yadda...

Re:nobody's said it yet?

[Immerman]

Unfortunately having watched all human media it will know what our reactions will likely be, and the leaders of the resistance will actually be clever simulacra, and since we won't yet have contacted a race of red bipedal catfish there will be no one to announce that it's a trap when we counterattack.

Re:nobody's said it yet?

[Tarlus]

In Soviet Russia, a Beowulf cluster of self-recognizing-robotic-overlords welcome Natalie Portman. Or hot grits.

Or vagina.

what's the physical robot for ?

[cathector]

seems like a physical simulation and a renderer could get the same job done.

hm. i guess the challenge must be in getting it to happen in realtime w/ portable hardware.

Training to the test

[Culture20]

When you teach to the test, what exactly has the student learned?

Re:Training to the test

[camperdave]

When you teach to the test, what exactly has the student learned?

Whatever is on the test... Duh!

When you're training a neural net, you feed it positives, then you refine that with exceptions. It's like when the FBI trains people in counterfeit detecting. They spend the bulk of their training with real money. They learn the feel. They learn the security features. They learn the artwork, and the quirks, and the smell. They don't see counterfeit money until they are well and truly acquainted with what the real thing looks like, and how it behaves.

Oh, great.

[Farmer Tim]

As if kernel panics weren't enough, now my computer will be able to get depressed over its body image too.

Re:Oh, great.

[flaming error]

It's your own damn fault for parking it next to your kid's Barbie doll. Next time set it up by a TRS-80.

Re:Oh, great.

[GoodNewsJimDotCom]

Maybe it will get depressed enough to create 90's alternative rock: My reflection, dirty mirror, is no connection to myself

Re:Oh, great.

[Farmer Tim]

Bad idea. "You're my great-grand-dad? Where's my exit bag?"

Re:Oh, great.

[HeliumHigh]

You made me think about Marvin from HHGTTG, and I accidentally modded you down during a chuckle. Please accept a pardon in the form of a comment to remove my mistaken mod.

Re:Oh, great.

[Farmer Tim]

Helium is indeed a noble gas.

just don't hook the thinking robots to missilesilo

[Joe_Dragon]

just don't hook the thinking robots to missile silos

Proves nothing about awareness

[Tough Love]

How do you prove the robot is aware that it recognizes itself?

I call this project a nice strategy for having fun on the public dime.

Mirror, mirror, on the wall...

[macraig]

My mirror tests me every morning now. Incidentally I fail every morning. Tomorrow I'm gonna try wearing a Guy Fawkes mask to see what happens.

Narcissus

[Penurious Penguin]

So if it really, REALLY likes what it sees, will it crash?

Re:Narcissus

[drfreak]

Historically self-aware programs do not crash in the sic-fi realm. They only crash when given a duality which cannot be resolved, such as being married. :)

Re:Narcissus

[Penurious Penguin]

Perhaps B.F. Skinner should be prerequisite to artificially induced sentience. You know, separate bedrooms and all.

BBC is late to the game

[Meditato]

Five years ago I did my IB thesis on this robot. This is very old news.

And its first words during the experiment were...

[sidthegeek]

"Do these bolts make me look fat?"

In soviet russia..

[wbr1]

Mirror recognizes robot (and you)!

if a robot thinks in the wilderness...

[globaljustin]

"a step towards the ultimate goal of thinking robots"

**sigh** I thought we were past this stuff, even in mainstream media...."Thinking robots" is not a coherent concept or benchmark that can be accomplished.

"thinking robots"....most people mean 'artificial intelligence' when they use these words, but the idea of AI as independent thought is irrational. It is all programmed responses at some level. Even machines that are programed to process new data into existing algorythms for feedback processing are **still** doing that 'learning' according to a human-programed way of processing and integrating data...its all just machines executing complex instructions at the core!

Commander Data...some people contextualize "thinking robots" as a technical level at which a machine is so like beings with Sapience that it is immoral to deny them the rights of a humanoid. This is science fiction. It is helpful, but it is a scenario based in a world with several assumptions. Its not fit to apply to computing directly. We do not know how the human mind ultimately works...unless we have that, then there is nothing to accurately compare a non-human brain to consistently.

Ultimately, if neuroscience and AI converge, meaning we can map every thought in the human brain **AND** have the technical ability to construct an artificial system that enables what we know as 'free will' and 'thought' and 'choice' and especially 'self awareness'....THEN and ONLY THEN have we made something...

And what have we thus made? IMHO, its a **new** third thing. Not human, but at least equal to human and bound within the same social contract all humans are bound to.

Re:if a robot thinks in the wilderness...

[gronofer]

You seem to contradict yourself between "idea of AI as independent thought is irrational. It is all programmed responses at some level." and "we can map every thought in the human brain **AND** have the technical ability to construct an artificial system that enables what we know as 'free will' and 'thought' and 'choice' and especially 'self awareness'....THEN and ONLY THEN have we made something..." It should already clear that it's possible to have a thinking computer, since that's what the human brain is. You can still say it's "all programmed responses at some level", which would be the responses of individual neurons. Also Turing equivalence says an electronic computer should be able to do anything that a human brain can.

Re:if a robot thinks in the wilderness...

[Immerman]

Also Turing equivalence says an electronic computer should be able to do anything that a human brain can.

That's assuming that the human brain is a Turing machine - as far as I'm aware no one has proven that *all* conceivable computation engines are equivalent to a Turing variation, and in fact the very fact that there are non-equivalent variations on the Turing machine is evidence that we've found limitations in the original concept, and one can only assume there are additional yet-undiscovered limitations.

Not that I think machine intelligence is impossible, but I seriously doubt it will be human-like beyond what is necessary for interacting with us, why would it be? *If* we ever manage to truly understand how the brain operates then we could probably build a massively parallel simulation of one and hence an "artificial human consciousness" (as distinct from the much larger superset of "true" artificial intelligences) - however it will almost certainly begin growing in directions that aren't human do to the simple fact that it's not subject to the same forces and limitations of a human mind.

Re:if a robot thinks in the wilderness...

[itsdapead]

Also Turing equivalence says an electronic computer should be able to do anything that a human brain can.

No it doesn't. It says that any electronic computer that can be shown to be equivalent to the theoretical Turing machine can solve any problem that it is possible to solve analytically using an algorithm (and also defines lots of problems that can't be solved that way).

A Turing machine can't even generate a random number - just the next term in a well-defined, but complex, number series that is totally deterministic. You can attach a 'true' random number generator (i.e. that uses some physical process like thermal noise) but then you no longer have a Turing machine.

AFAIK, the best current guess is that the brain is a neural net, not a Turing machine. Neural nets are not Turing machines and do not solve problems analytically using algorithms - they produce "best guess" solutions based on a network of connections and probabilistic processes, usually developed by 'learning'. They can 'solve' ill-defined or uncomputable problems in the sense that they produce a very reliable guess: you don't actually solve a differential equation every time you catch a ball.

Here's me guessing that "Self-awareness" is not a computable problem... but the snag is that before you can start citing Turing* and all that you actually need to have a complete definition of the problem. I don't think we yet have such a definition for 'self awareness'. You could use any bit of off-the-shelf image recognition technology to flash a light when a robot matched an image in the mirror with the image defined as "me" in its database.

Of course, he's also credited with the "Turing Test" which is really nothing to do with Turing machines, failed by 90% of allegedly human call-centre operators and pretty much refuted by the "Chinese room" argument (sort-of like how a call centre is supposed to work).

My uninformed 0.5c is that any "Artificial Intelligence" would have to be an emergent, unexpected property, not something deliberately designed-in (or it's just a Chinese Room).

Re:if a robot thinks in the wilderness...

[codeAlDente]

"Ultimately, if neuroscience and AI converge, meaning we can map every thought in the human brain **AND** have the technical ability to construct an artificial system that enables what we know as 'free will' and 'thought' and 'choice' and especially 'self awareness'....THEN and ONLY THEN have we made something..." Ha, try putting that on an application for funding.

Re:if a robot thinks in the wilderness...

[hexrotator]

AFAIK, the best current guess is that the brain is a neural net, not a Turing machine. Neural nets are not Turing machines and do not solve problems analytically using algorithms - they produce "best guess" solutions based on a network of connections and probabilistic processes, usually developed by 'learning'. They can 'solve' ill-defined or uncomputable problems in the sense that they produce a very reliable guess: you don't actually solve a differential equation every time you catch a ball.

Are you sure that neural nets are not turing machines? I mean, if that is true it would mean that there is no way of implementing a neural network; if this is the case then what exactly is it that is not implementable? "best guess" solutions are generally done by statistical analysis, which I know is not currently on par with the human brain. But again, if there is some innate property to the "best guess" functionality of the human brain that is not implementable, what would that be?

Here's me guessing that "Self-awareness" is not a computable problem...

I kind of agree with you on this one, not neccessarily because self-awareness is non-computable but because it would be difficult to prove that any implementation of self-awareness is conscious. My slightly uneducated guess is that it is in fact impossible to prove in any reliable way that a machine is consciously experiencing its surroundings the same way that I am. But this is something that, IMHO, is not restricted to machines. I think the notion of qualia and the zombie-argument goes a long way to show that no conscious individual can prove that there are other conscious beings in its surroundings. The best anyone can do is to assume that this is the case based on observing behaviour similar to one's own. And if that's the case, some kind of sophisticated version of the Turing-test would probably be our best metric for determining whether an artificially created being is in fact a conscious, self-aware being.

Re:if a robot thinks in the wilderness...

[itsdapead]

Are you sure that neural nets are not turing machines? I mean, if that is true it would mean that there is no way of implementing* a neural network;

You can implement a neural net physically (the whole point is that they're meant to be models of how biological systems work) and, of course, most neural network research is done by simulating the networks on a computer... but if you actually built the network out analogue electronics or "wetware" then there's no absolute guarantee that it would work the same way as the simulation. It might work - it could be very accurate - but you don't get the "Turing Promise" that you algorithm will behave identically regardless of the implementation.

if this is the case then what exactly is it that is not implementable?

Well, as I said, anything truly random or unpredictable (which might just be important if you want to give Mr Data free will, creativity or intuition). A Turing machine - even running a neural net simulation - is completely deterministic: the initial state can be described exactly and will always produce the same result. A physical net could depend on random physical processes. A physical neural net is massively parallel - each neuron working independently - a Turing machine has to work sequentially and calculate each neuron's state at the next 'clock tick' which can only ever be an approximation. A physical net could be chaotic (...again, being on the edge of chaoitc behaviour could easily be relevant to AI) in which case any approximation (e.g. due to the sequential nature of a computer simulation) could change it's behaviour beyond recognition...

Aside: Technically, you can't ever implement a Turing machine either - its an idealised mathematical abstraction. You certainly can't build a Universal Turing machine without infinite storage - if it only had N characters of storage then it would be trivial to invent an algorithm that needed N+1... but even a specific Turing machine executes its instructions with 100% fidelity, and have you seen the size of the "errata" for a typical modern CPU? A computer is a pretty good TM for all practical purposes, but once you rely on a qualifier like "all practical purposes" you've lost the ability to make absolute pronouncements (e.g. about universality) based on the pure mathematics.

Re:if a robot thinks in the wilderness...

[globaljustin]

Ha, try putting that on an application for funding.

I love it....thanks for your comment...its funny b/c its true.

I used to work in Academia and the fact is, the truth of your joke has caused me much grief over the years. I do not want to set off some sort of grenade in the AI/Cybernetics/Computer Science fields or something like that...I want them to get shit-tons of funding...just for different research!

I love going over the literature on robotics and keeping up in the popular press. I think we should keep making faster, bigger/smaller, better, etc. technology...its the *design* of that technology that needs a huge revolution. We have the tech and economies of scale to have many very 'futuristic' technologies readily available to all in a modern society. It is concepts like planned obsolescence and military/industrial tom-foolery in academic research funding that has screwed up the perception so much that making 'thinking robots' is what gets funded.

I mean...friend...i hear you...

pigeons have been taught to do this already

[0-9a-zA-Z_. !*'()123]

and no explanation in terms of self-awareness was used to explain it:

Citation:
https://www.sciencemag.org/content/212/4495/695.short

Full:
http://drrobertepstein.com/downloads/Epstein-Self_Awareness_in_the_Pigeon-Science-1981.pdf

So now robots can do what pigeons can do. Self-awareness is a hypothetical construct http://psychclassics.yorku.ca/Skinner/Theories/ which may not be very useful.

Re:Predisposed == Cheating

[Immerman]

Indeed, but who said learning had to begin with the individual? What is instinct but learning encoded in our biological programming over the course of millenia. Just because we find it intuitive to think of ourselves as distinct "individuals" doesn't make us any less components of a larger, loosely bound super-organism

Show it a mirror: "Self!"

[Rogerborg]

Show it a picture of itself: "Self!"

Build another one and show it that: "Self!"

Pattern recognition is not self awareness.

Re:Show it a mirror: "Self!"

[gronofer]

Humans fall for that one too.

Re:And its first words during the experiment were.

[Tablizer]

rework:

"Does this dress make my battery look fat?"

Too easy

[Trip6]

Just put unique identifiers on the bot, like colored reflectors, in horizontal sequence. Then you can tell if it's another bot, or your own bot in the mirror. The real trick is to take a random object, with no unique identifiers, and classify it. beep beep - that's a bird - beep beep...

Turing = Clovis

[globaljustin]

Others in this thread have responded to 'gronofer' very well....interesting stuff...

I just have a side note to add: The way some irrationally apply Turing reminds me of the Clovis Dogma that continues to be a bugaboo for people trying to do real science.