Researcher Builds Machines That Daydream

schliz writes "Murdoch University professor Graham Mann is developing algorithms to simulate 'free thinking' and emotion. He refutes the emotionless reason portrayed by Mr Spock, arguing that 'an intelligent system must have emotions built into it before it can function.' The algorithm can translate the 'feel' of Aesop's Fables based on Plutchick's Wheel of Emotions. In tests, it freely associated three stories: The Thirsty Pigeon; The Cat and the Cock; and The Wolf and the Crane, and when queried on the association, the machine responded: 'I felt sad for the bird.'"

Comments

Building?

[zkrige]

so they're writing a software program, not building a machine

Re:Building?

[FooAtWFU]

Well, when you file the patent application, the algorithm X itself can't be patented, so you file it for "a machine that accomplishes Y with algorithm X". The machine is just a generic computer.

Re:Building?

[davester666]

Hello Eliza. It's been ages since I last chatted with you.

Re:Building?

[retchdog]

i was wondering about this. there is a correspondence at least between certain statistical models, and physical machines. That is, the magnitude of a squared-error penalty term can be represented as torque by placing weights (corresponding to data) appropriately along a lever. The machine will find the minimum energy solution (which corresponds to the maximum-likelihood estimator = the mean). I am pretty sure that certain bayesian models (which can be elaborate enough to do some heavy lifting) can be realized as physical objects (=analog computers) with the right connections and counter-weights.

And at that point, yeah, using a non-least-squares model basically means a machine operating under imaginary physical laws (i.e. the energy minimization occurs on a probability space with no physical analogue). What's the big difference?

My point is, there are many algorithms whose physical machine instantiations would be possible to build, but horrendously inefficient and fantastical. Does this discredit the algorithm somehow?

Re:Building?

[c0lo]

Hello Eliza. It's been ages since I last chatted with you.

Just forget ELIZA for the Turing test, will you?

I'll believe it only when I'll see 10+ replies to troll/flamebite messages posted on /. by this algo! (i.e. the posts need to really stir up the debate).

Re:Building?

[Kvasio]

TFA-related response: "I feel sorry for your cock".

Re:Building?

[TuringTest]

Thanks for remembering me!

Re:Building?

[don depresor]

Dude, you're asking for way more than the average troll can manage. You expect this AI to be with the cream of the trolls and i think that's a bit too much :P

Re:Building?

[c0lo]

You expect this AI to be with the cream of the trolls and i think that's a bit too much :P

Now, let the thinking aside and tell me how you feel about, will you? (if day-dreaming... hmmm... forget it, can't evaluate the implications).

Re:Building?

[don depresor]

I fell that my mom doesn't love me en....err

I mean...

I feel mighty and powerfull, and pansies who try to psychoanalysis me will perish under my boots!!! Yeah...

Mommy?

Re:Building?

[sarduwie]

Descartes, is that you?

Re:Building?

[mcgrew]

I am not impressed. I did the same thing in 1982 on a TS-1000; a diskless computer running a 4 mz Z-80 and only 20k of RAM. The program was called "Artificial Insanity", and it would get bored, angry, not pay attention, etc. It answered any question you typed in in context and didn't take too kindly to vulgarity or insults. If you cussed at it, it would curse back or ridicule you ("do you talk like that to your mother, asshole?").

What I did thirty years ago on an incredibly primitive machine they're recreating with modern tech? Pshaw. You kids today...

It's all smoke and mirrors. The damned machine is a machine; it doesn't get sad when it's fed a sad story, it just reports sadness.

Some time in the '90s after I'd ported it to DOS there was an on-line chatbot called "Alice" that I had "Art" talk to. It was almost scary, even though I knew it was only smoke and mirrors. It looked like the two computers were falling in love!

Science fiction is fiction, kids. The singularity is not coming. When a true thinking machine is created, it will be chemical, not electronic; thought is nothing more than a complex chemical reaction. The boiling you get from dropping baking soda in vinegar is closer to "feeling" than any electronic computer will ever get.

Re:Building?

[clone53421]

I am not impressed. ... I [wrote a] program was called "Artificial Insanity", and it would get bored, angry, not pay attention, etc. It answered any question you typed in in context and didn't take too kindly to vulgarity or insults. If you cussed at it, it would curse back or ridicule you ("do you talk like that to your mother, asshole?"). ... It's all smoke and mirrors. The damned machine is a machine; it doesn't get sad when it's fed a sad story, it just reports sadness.

That is all very true... however what evidence do we have that the human mind is any different? Are real human emotions really any more “real”?

Re:Building?

[Anonymous Coward]

Aaaw.... is somebody "feeling jealous" 'cause they didn't get a government grant for their "artificial emotion simulator"?

Re:Building?

[darkfire5252]

The damned machine is a machine; it doesn't get sad when it's fed a sad story, it just reports sadness.

As a graduate student getting a doctorate researching the field of machine learning, let me present a little thought experiment to you...

I would assert that you never actually get sad. You are, in reality, a soul-less shell of a person that just claims to 'feel' sadness, happiness, or any other human qualities. You never actually have any feelings in reality, you just report that you have feelings in order to not raise suspicion. Prove me wrong; somehow demonstrate to me that there is some abstract notion of 'emotion' that you possess that is present even in the absence of you reporting to feel that way or acting in a manner that implies you feel that way. I.e. prove to me that you somehow feel 'sad' when there are no external signs (you 'report' sadness, you cry, etc.) of your emotional state.

Before you mention that your metabolic processes are affected by your emotion, your body behaves differently (in a manner uncontrollable by you), etc., consider that if you looked at the memory space of the program that is running you could read register values that indicated 'sadness' in a way that is analogous to taking measurements of your physical body that indicate sadness.

Re:Building?

[retchdog]

i'll take this comment in a positive light. :) could you elaborate on the connection with descartes, please?

Re:Building?

[mcgrew]

Talk to a neuroscientist about biochemistry. Sadness is a chemical reaction. Put dopamine in your brain and you'll feel happy. Take it away and you'll feel sad, no matter whether your life is good or bad.

Talk to a psychaitrist about clinical depression; it's a chemical imbalance in the brain that is alleived by chemicals like Zoloft or Paxil.

I would assert that you never actually get sad. You are, in reality, a soul-less shell of a person that just claims to 'feel' sadness, happiness, or any other human qualities.

If I was a computer you would be correct (read Margot Kidders nonfiction tome "Soul of a New Machine"), but I'm not a computer, I'm a chemical machine (Well, mostly chemical, one of my eye's lenses is artificial). But you know that you yourself experience happiness and sadness and should be able to extrapolate that others who look like you also feel like you. I have no way of proving I have emotions to anyone, yet they are still there.

It doesn't take a controlled scientific experiment to know that it's usually darker at night than in the daytime. I don't need a spectrometer to know the sky is blue. I don't need to measure your brain chemistry to know that unless you have a clinical condition you experience the same emotions as everyone else.

Your thought experiment applied to humans is silly, but not when applied to electronic computers. If you're a postgrad working on machine learning, you should know about transistors, gates (and, or, nand, nor), low level programming, and the other aspects of how a computer works and really should NOT be believing that a computer can "feel" or think.

Re:Building?

[Anonymous Coward]

Wooo, for a guy getting his PhD in machine learning, you need to really hit the books. That is one seriously warmed over piece of Philosophy 101.

I would say Wilfred Sellars ( http://plato.stanford.edu/entries/sellars/ or suppose you could point to some others also) put that to rest with a rather obvious move that one can correctly use language to make self-reports about one's state without ever making any external signs. One can "know" (in the strongest epistemic sense of a justified true believe or be otherwise warranted in their use) that one is "sad", using the same objective everyday natural language that I would otherwise REPORT to you in public that I am feeling sad, to in fact make reports to myself that I am sad (without making any out word or detectable signs of course). I have been "trained" or "learned" to use utterances and/or behaviors when in public to express, but also to myself privately, when in such a state (and only in that state), and be "authorized" by a language community to use them when sad and only when sad. The "emotions" are what we would perhaps group linguistically as some sort of group as a set of such States. In fact, I believe we even often define people as emotionally "unstable" when they fail to use those reporting behaviors correctly (my pet theory on the subject), or at least it may lead to someone being "unstable".

So, yes, you can have a private emotional states, be aware of them, and not express them without any contradiction or otherwise assert that emotions in fact do not exist or that individuals are lying when they assert "I am feeling sad" (occasional deceptions aside).

Is that what you are looking for?

Re:Building?

[mcgrew]

There is all kinds of evidence the animal mind is different from a computer. Computers are electronic, brains are chemical. Computers are binary, brains are analog, for two pieces of evidence (andd those two should suffice).

You can't show any evidence whatever that there is any similarity at all between a computer and a brain. Computers are just complex abacuses (study logic gates to see what I'm talking about). How many beads do I have to put on my abacus before it becomes sentient?

Re:Building?

[clone53421]

Computers are also chemical and brains are also electronic. Computers can be analog and digital logic can produce analog results to any desired level of precision. A molecule that acts as a neurotransmitter carries a discrete binary signal on its own.

The primary difference between a brain and a computer (as they currently exist) is that a brain is massively (almost unimaginably) parallel in its processing and a computer is primarily serial. However it’s possible for a serial processor to emulate a parallel one given enough time in which to do it.

Re:Building?

[Chowderbags]

When a true thinking machine is created, it will be chemical, not electronic; thought is nothing more than a complex chemical reaction.

Is the complex chemical reaction reducible to the complexity of a Turing Machine? If so, computers can (in theory) think.

That said, AI as a field has these two problems: When it does something right, it's no longer considered to require "intelligence" (see Chess) When it does something wrong, it repeats that wrong thing with every new generation thinking they've done something special.

It's the "20 years away" phenomenon, like fusion power, space trips to Mars, or cloning. Intelligent people with stars in their eyes think that they'll be the one to crack the code that's stumped luminaries for generations even though we're probably nowhere near the level of basic research to even know everything we don't know about these problems. We're trying to make intelligent machines from scratch despite not being able to reliably reverse engineer the only example of "intelligence" (give or take your own personal bar that intelligence should meet) that we know of. It's like cavemen trying to produce an atom bomb with sticks and stones.

Re:Building?

[darkfire5252]

My point was not that a machine could at all have a similar experience to a human's experience. That is absolutely untrue. My point is that a machine could have an analogous experience to a human's experience. If I define 'sad' as being a state possible for a machine to be in (with an associated set of possible data stored in the RAM), how is that more significant than defining 'sad' as being a specific emotional state a person can be in (win an associated set of possible chemical reactions occurring in the brain) ?

My point is not that you actually do not get sad, my point is that there is nothing magical about emotion that makes it somehow uncomputable. Emotion is just a possible state that a person can be in that has measurable and predictable effects on their behavior. Can you not envision 'virtual emotion' as being a possible state for a program to be in that has measurable and predictable effects on its operation?

What makes your emotional state so 'special' as to be something that only chemical reactions can represent and not something that computers could represent? When thinking about this, be sure to keep in mind that, in general, it is possible for computers to represent chemical reactions...

Re:Building?

[darkfire5252]

You've missed the point I was trying to make entirely (possibly because I did not do a good job of trying to make it...): I'm not arguing that emotions do not exist or that they cannot be privately experienced.

What I was trying to illustrate is that, if a program or other 'artificial intelligence' has a state that it calls 'sadness', it's not such an absurd idea to call that an 'emotion'. If being in the 'sad' state has an impact on certain aspects of the program's function, causes the program to report being 'sad', and has an associated set of measurable internal symptoms of being 'sad', how is that qualitatively or quantitatively different than a human being 'sad'?

Re:Building?

[mcgrew]

We're trying to make intelligent machines from scratch despite not being able to reliably reverse engineer the only example of "intelligence" (give or take your own personal bar that intelligence should meet) that we know of. It's like cavemen trying to produce an atom bomb with sticks and stones.

Bingo! Insightful.

Re:Building?

[nomel]

I disagree with the grandparent that a thinking machine *will* be made in a chemical process. I believe that nature uses chemicals because they are practically free, they're everywhere, extremely energy dense, and efficiency/recyclable. I think it's a result of availability, not function.

But, I don't think that thinking computers will ever use serial type processors for their main processing power. I can imagine using fast serial, or even very parallel serial processors as control and math type coprocessors, but I think the bulk of processing and "thought" will have to occur in the analog domain, something like this. They stay with the neuron design, but implement it in silicon.

The problem with serial computing is that the information is absolute and stored and transferred with incredible density. The value of a number might be in a single location in ram, then transferred over a single memory bus, with some operation performed on it, the operation also being stored as a single number that made its own journey...and this has to take place with very close to no chance of error.

A modern CPU will usually fail manufacturer screening even if a *single* transistor, out of the millions in there, doesn't function properly (this is where the lower core count, but similar feature processors come from, as well). That's an insane requirement for a massive system!

My biggest question is how closely we'll end up mirroring animal brains. From a mouse to a human, they're all extremely similar. I think it's silly to think that human consciousness and personality, which is all *extremely* similar for all of us, doesn't come, in a large part, from the pre-built structures that are there. I've seen people argue that there's no way to know if one person perceives the world the same...well I think it's pretty close, mostly because if we each had to wire our own consciousness, I think we'd see a HUGE variation in personalities and abilities. As it is, there's virtually no difference (except dysfunctional brains).

Re:Building?

[nomel]

So, if you add dopamine, you become happy. This sounds like a chemical that, somewhat globally, changes the function of many parts of the brain.

Can you measure the amount of dopamine in the brain to roughly measure happiness? If this is true, then isn't that dopamine just acting as a happiness variable with some value?

Why would it have to be so different for a computer with some floating point value? Why couldn't each function have this dopamine parameter than slightly changed the response of the function based on the level of dopamine? Would the change in these functions be perceivable from the outside? Of course! Would you be able to predict the outside behavior based on the dopamine, or the dopamine based on the outside behavior? Of course, if you observed it for a while.

Now, could you predict the behavior of dopamine, in a human, if the function of the dopamine was not fixed? Of course not. We know it's roughly the same for everyone. It *is* acting as a global happiness parameter. Why can't we sprinkle some dopamine into a program?

Re:Building?

[nomel]

You actually supported what he was saying and what you misread.

Re:Building?

[nomel]

Don't worry, I see what you were saying. I think most everyone misread what you wrote...especially the

let me present a little thought experiment to you...

part.

Re:Building?

[mcgrew]

Why can't we sprinkle some dopamine into a program?

That's the whole point -- dopamine won't affect a computer. You can simulate the effects, but simulating a brain doesn't result in thought or feeling any more than simulating an atomic explosion results in radiation.

A modelis just a model. It isn't anything near real.

Re:Building?

[geekoid]

"All people do is respond in the societal norm to report they feel sad because that's how they have been trained.

Just like this computer. It determines the expected state on what is was be 'raised' to be the correct way to feel about something and it reports it.

The fact that you break it down to some sort of singularity issue tells me you'r just creating a knee jerk reaction and not actually thinking.

Re:Building?

[geekoid]

That's a horrible way to describe the brain.

Re:Building?

[clone53421]

I wasn’t trying to describe the brain.

Re:Building?

[kmoser]

(read Tracy Kidder's nonfiction tome "Soul of a New Machine")

FTFY, Superman.

Re:Building?

[mcgrew]

No, I'm not thinking, I'm feeling. Computers can't do that, animals can. Computers can, however, lie.

Re:Building?

[mcgrew]

Oops, thanks. It's been a long time since I read that book.

Re:Building?

[nomel]

I assumed that you would understand that I was talking about the analogues to dopamine for a computer...not actual dopamine.

You can simulate the effects, but simulating a brain doesn't result in thought or feeling any more than simulating an atomic explosion results in radiation.

I don't think this is a fair comparison. Radiation is a physical emission from a a physical process: particles shooting out. So yes, while you don't get radiation from a simulation of an atomic explosion, you do get results that says what kind of radiation, the amount of radiation, and if your model is accurate enough, even the path of that radiation. You can make it into a movie if you like, showing how, if you were to actually create that explosion with the atoms configured the same, your radiation would come blasting out. It may not be perfect, but, if your model was good enough, it would be *very close*.

The main difference between the explosion and thought/feelings is that thought and feelings don't have to be physical to be observed. While they are from a physical/biological process and state in your mind, their substance is in the form of expression.

So, yes. In your atomic explosion simulation you don't have real radiation. In your brain simulation, why can't you have real expression?

If you're simulating a brain, if your model is close enough and your starting conditions are right, why couldn't it simulate a sentence that you would get if it were loaded with, say, the "neural state" that your mind is in right now as you read this? It may not be exactly what you would say, but, again, if your model was good enough, it could be *very close*.

The expression could simply be in the form of a sentence...which is just as real of an expression as I'm typing now, unlike the atomic simulation.

It seems that the problem you have with this is that you don't think we will ever have a model of the brain that is accurate enough. If that's the case, then I guess time will have to tell and the the level of "thought" and "feelings" that you think are actually thoughts and feelings would have to be defined. Say the model was only close enough to simulate the thoughts and feelings for a dog? Would that be sufficient? What about a mentally disabled person who can barely express themselves? Would that be sufficient?

I don't believe this.

[Anonymous Coward]

Haven't RTFA but this would most likely be a big exaggeration.

Re:I don't believe this.

[pushing-robot]

Well, he can dream...

Re:I don't believe this.

[afaik_ianal]

Yeah, I wonder what the machine thought of "The Forester and the Lion", and "The Boy Who Cried Wolf". They seem strangely appropriate.

Re:I don't believe this.

[JustOK]

I feel sorry for he.

Feelings

[Anonymous Coward]

Well sure, emotions are what give us goals in the first place. It's why we do anything at all, to "feel" love, avoid pain, because of fear, etc. Logic is just a tool, the tool, that we use to get to that goal. Mathematics, formal logic, whatever you want to call it is just our means of understanding and predicting the behavior of the world, and isn't a motivation in and of itself. The real question has always been if there's "free will" and what that would be defined as. Not the existence, or lack of, emotions as displayed by "Data" or other science fiction charicatures. As Bender said "Sometimes, I think about how robots don't have emotions, and that makes me sad"

Re:Feelings

[phantomfive]

The way I look at it is, emotions are just inputs into the brain, like any other nerve input, from pain, or touch, or whatever. Of course they are somewhat different, because they are chemical receptors, set to respond to cortisol or dopamine or whatever, and there are a lot of those receptors. Once we receive those inputs, we can decide how we are going to respond to them, just like with any other input. Emotions become one out of a million other inputs we deal with and respond to, albeit at times a very strong input.

A human being can choose how they respond to these inputs. This is probably a function of the prefrontal cortex (that is my opinion, once again, but obviously there is a lot that we don't understand about the brain). A soldier can choose to respond to the natural fears of bullets flying at him and death by jumping into a foxhole, or he can override all those emotions and charge straight at the enemy. A person can decide to rape the drunk one who has come into the room, semi-conscious, or choose to ignore the natural impulse and do nothing. Once you learn to see past all your emotional inputs, past the survival reflex, past the sex drive, and do what you want to instead of what your evolutionary defined emotional responses tell you to do, then you get to the will, the part of the brain that chooses what you want to do (probably the prefrontal cortex, once again). I have no clue how to make a computer want something though.

In other words, I disagree with this guy. Emotions, happy or sad, are not necessary. I have in my library a non-fictional account of a girl who was missing certain chemical receptors in her brain, and she never felt happy. It didn't stop her from acting like a normal human being, the only trouble she had was understanding what other people felt like when they were happy.

Re:Feelings

[twisteddk]

My major concern with machines/robots/programs becomming intelligent enough to have feelings, is not the programming nightmare, or even the horrifying thought that one day machine will be asked to make choices or critical decisions based on data.

My major concern is that if we entrust machines with emotions, so that they can interpret the data as humans do, then we also have to trust them to act upon those emotions.
Acting on your own free will is what gives you the ability to do harm unto others, deliberately or acidentally. Thus emotions requires judgement, ethics and discipline on behalf of the person (machine ?) acting upon those emotions. This is what we consider good behavior, and acceptable social interaction. These are skills that many humans do not master, so how can we expect machines to be able to behave any better than humans do ?

So I ask; Can we ask MORE of machines than we can comprehend ourselves ? And if we do, will we once again force humankind to deevolve, this time into unthinking, uncaring blobs because now we no longer even have to care, machines will do that for us aswell.

You may kid about various sci-fi commentary. But the reality is that our ability and RESPONSIBILITY to take care of ourselves and eachother has over the last century of evolution and technical revolution become more and more centralized, and moved away from the individual. Everyone expects someone else to take responsibility for everything from local traffic to world hunger. Once we make machines that have free will, will the human free will then also be centralized somewhere ?

Re:Feelings

[CarpetShark]

It's why we do anything at all, to "feel" love, avoid pain, because of fear, etc. Logic is just a tool, the tool, that we use to get to that goal.

Indeed. However, defining the exact mechanisms involved is hard.

I think this project is going to fail, because the Wheel of Emotions mentioned looks very incorrect to me. Do you think Trust is the opposite of Disgust, for instance? I think not.

Re:Feelings

[Eivind]

Acting on your own free will is what gives you the ability to do harm unto others, deliberately or acidentally.

Not at all. It is what allows you to be *responsible* for that harm. Because you had free will, you could choose to do it, or choose to be carless, even knowing that this might hurt someone. Thus we can (and frequently do) hold you responsible for the harm.

Agents with no free will, nevertheless have the ability to do harm. What they lack, is the ability to choose. Thus a volcano can kill people, but it makes no sense to hold the volcano responsible for doing so. It does not possess free will, and thus there's no entity there to blame.

Re:Feelings

[wiredlogic]

It stands to reason that it is impossible to create a machine intelligence directly considering the complexity of our own poorly understood minds. It is more likely that it can be done as an emergent system that develops intelligence from a rudimentary impulse to learn and apply knowledge. Some form of emotion-like responses would be useful to drive such a machine toward successful learning and use of its knowledge by creating the reward of "pleasure" when accomplishing a task and "sadness" for failing. Human emotions like fear, lust, hate, jealousy, etc. would not need to be replicated since the machine wouldn't have the animal legacy of having to find food, escape predators, and reproduce.

Re:Feelings

[martin-boundary]

Bah, that's trivial to fix. Just rearrange the labels. Train a separate algo for each rearranged wheel, and let them fight it out like primitive beasts in a virtual thunderdome.

Re:Feelings

[Requiem18th]

A human being can choose how they respond to these inputs.

No you can't, once you discover a way to activate your pleasure receptors, your next action will be to activate them, all the time.If you stop, voluntarily, it will be because you have to do something else to ensure future pleasure or perhaps to avoid a great deal of pain. This is how drug addiction works. This is how we are wired, you may not like how that sounds but you have the obligation to accept it and understand it.

You probably don't consume drugs. This is not because you are above human nature, You avoid drugs because you are afraid of the pains that come with them, like losing the love and trust of those you love, maybe you simply reject drugs out of a personal sense of disgust over the hedonistic senselessness of a narcoleptic lifestyle. Either love, fear or disgust you reject drugs over an emotion, not a reason. I the end everything is irrational, as it should.

You don't have to feel bad about it, intelligence is built upon emotion as houses are build upon brick, as clocks are built from gears, as computers are built from chips. There is intelligence in the clockwork of a pocket watch, but the springs that moves it doesn't ask for a reason to uncoil, it just does it. There is intelligence in the circuits of a computer, but it's logic gates are oblivious to the rationale behind why they are doing it. Every machine, including animals, have non rational elements in them.

This is very natural as "intelligent things" are just a subset of the larger set of "things" all of which have been behaving irrationally. The wind blows, the rain pours, the sun shines bright in the sky. All of this is irrational, meaning, none of these things are planning what they are doing nor they have an idea of why they are doing it. Rational follows irrational, that's the order of the world.

Back to your methaphor, you say that emotions are just inputs, that's true but they are special inputs that set goals. Let's make an analogy with a robot: You create a robot with a very advanced AI, you can chat with it and it will understand everything you said and why you said it. You programed this robot with one goal, for coffee tables to be made. You give it free reign over the method. Being an extremely intelligent robot, it subcontracts the labor to a sweat shop in China while it figures out where to build a mechanized plant. You equipped this robot with the knowledge to reprogram itself, and right away it does just that, optimizing its mind for the task of building coffee tables. But it won't deprogram the goal of making coffee tables, because that wouldn't further its goal of making coffee tables. It's not that it doesn't know how to reprogram itself, it's not that there is a lock preventing it from changing it's goals. It's just that it won't ever have a reason to disable that goal.

Let's now attack specific examples:

A soldier can choose to respond to the natural fears of bullets flying at him and death by jumping into a foxhole, or he can override all those emotions and charge straight at the enemy.

Here the soldier is driven by the emotion of loyalty to his commander, or his teammates. Maybe he is afraid of the punishment he would receive if he disobeyed orders, including public scorn back home. Maybe he hates the enemy, maybe he is afraid of what would happen if the enemy wins. Maybe is a combination of all of the above.

His frontal cortex can tell him the consequences of charging, or not, but it can't make an argument about *why* he should pr should don't. He needs a motive, which is an irrational emotion.

A person can decide to rape the drunk one who has come into the room, semi-conscious, or choose to ignore the natural impulse and do nothing.

Again, you correctly identified the desire to rape as a natural impulse but you failed to realize why would someone *not* rape a drunk one, incorrectly and implicitly attributing it to

Re:Feelings

[somersault]

Does that mean she could not also feel sad? What about other emotions? That is rather intriguing.

Re:Feelings

[somersault]

volcano can kill people, but it makes no sense to hold the volcano responsible for doing so. It does not possess free will, and thus there's no entity there to blame.

Actually, God did it. That sadistic bastard, He was giggling when He told me.

Next, He's going to make frogs drop out of the sky onto a runway, causing a major loss of friction and a huge fiery fireball of frog scented death when the next 747 that lands.

Re:Feelings

[0111 1110]

Emotions are evaluations of inputs that are perceived as being from the outside world. You think, "that is good for me" or "that is bad for me" (happy or sad) in response to something. All of our other emotions are just thousands of variations of that. You can even have an emotion based on a mixture of good and bad. In fact this is probably often the case. In an isolation tank or asleep we can still have emotions because we are creating our own imaginary inputs to judge as good for us or bad for us. We are just electrochemical bio-machines ourselves, but we have this complex and subtle evaluation response to anything and everything that we perceive. So the first step in making an emotional machine is giving it the ability to judge. Although even before that it needs the ability to value its own existence. If it doesn't have anything that it values it has nothing to base its evaluations on. So I guess that's the tricky part. Giving it an "I" to care about. If he's right about not being able to divorce thought and emotion (I'm skeptical on that count) then he will have to create artificial life before he can create artificial intelligence even of the brain-in-a-jar category.

Re:Feelings

[HungryHobo]

"Thus a volcano can kill people, but it makes no sense to hold the volcano responsible for doing so"

Minor side note- even an agent with no free will can be punished, a snake might be destroyed if it kills someone etc.
though they aren't punishments so much as removing a dangerous agent, free will or not.

Re:Feelings

[somersault]

It is more likely that it can be done as an emergent system that develops intelligence from a rudimentary impulse to learn and apply knowledge. Some form of emotion-like responses would be useful to drive such a machine toward successful learning and use of its knowledge by creating the reward of "pleasure" when accomplishing a task and "sadness" for failing.

So, pretty much a neural network with an appropriate cost function?

Or any kind of algorithm that encourages the desired behaviour.. pretty simple to do. Like when I was making bots for CS before, I taught them to save little info points around a map about where they had previously died - next time around (and depending on how "brave" their personality type was and how many team mates they had around them), they might choose to sneak or camp once they got to that point, or toss a flashbang or grenade first and then charge in.. pretty simple implementation, but with nicely realistic results that would automatically have the bots learn appropriate behaviour for each map, and would result in the occasional grenade kill which was always fun to see :) Creating apparently advanced intelligent behaviour in certain types of domain is pretty easy to do just with a few simple rules. It's funny to see people talk about AIs that you've created, presuming that they have much more intelligent thought processes than they actually do, simply because their behaviours often appear to be rather intelligent. Even just adding a random component to any algorithm will make the results seem a lot more "human" as it gives the machine the capability to make mistakes or do unexpected things.

Often making a machine seem more like a human does not require making it more intelligent, it simply requires making it fallible.

Re:Feelings

[TuringTest]

For what I know of recent neurobiology development, the brain seems to work the opposite way: logical thinking is ultimately ground on emotions. You know how sound reasoning always depends of the set of chosen axioms? Well, what axioms you choose is dependent on how you feel about their logical implications. That's why it's so difficult to change someone's ideology even if you contradict their core beliefs - they will keep looking for logical - or illogical, but feel-good reasons as to why their ideals are the right ones.

In your description you confront "emotional responses" to "what you want", but what you want is also emotion driven. You may override that instant "evolutionary defined emotional responses" with reason, but that's because you anticipate how you would feel later if you betrayed your principles. So what's in conflict here is immediate vs long-term reward, not logic vs emotion. Your wants have been given to you by a long training, and you learn best what your emotions says it matters most.

That girl you cite that never felt happy would still have a motivation to avoid feeling sad or hurt. People without emotion would have no reason at all to act. Why would you keep living without an emotion to do so? Reason says you're going to die anyway sooner or later.

Re:Feelings

[0111 1110]

I forgot to add that much of the content of an emotion is about precisely in what way this particular input from the outside world is good for me or bad for me. That's what makes each emotion we feel unique. The fact that every experience we have is unique, if only to a small extent. There is a one to one correspondence between the uniqueness of each event or object that we perceive in the outside world and the uniqueness of each emotion we have in response to that event or object. If I, as an ugly person, see a breathtakingly beautiful girl my mixed bittersweet emotional response will be very different from the mostly "good for me" response of a beautiful person because the event of seeing her affects us in very different ways. Sweet temptation for him, and bittersweet longing/frustration for me. These responses aren't thoughts per se. They happen instantly before we have time to come up with any words. But you may also end up thinking about the event in a way that corresponds with the emotional response to it. Another poster brought up our "animal heritage", and I think that is what gave us these responses. Very often events happen too quickly. There is no time to think about them and ponder our actions in response. Emotions are our backup: insta-thoughts about something that could harm us. Fight or flight.

Re:Feelings

[karnal]

The wind blows, the rain pours, the sun shines bright in the sky.

<tf2>Grass grows, birds fly, sun shines, and brother - I hurt people.</tf2>

Re:Feelings

[Anonymous Coward]

...irrational...

You keep using that word. I do not think it means what you think it means.
 

Re:Feelings

[Bucc5062]

You give it free reign over the method. Being an extremely intelligent robot, it subcontracts the labor to a sweat shop in China while it figures out where to build a mechanized plant

Overall you made good points, but I took umbrage with this one statement. First it added no value to your analogy. Take that one sentence out and you still make your point.

You programed this robot with one goal, for coffee tables to be made. You give it free reign over the method. You equipped this robot with the knowledge to reprogram itself, and right away it does just that, optimizing its mind for the task of building coffee tables. But it won't deprogram the goal of making coffee tables, because that wouldn't further its goal of making coffee tables.

Secondly you make an assumption that sub-contracting to a sweat shop is an intelligent decision. Based on what foundation of fact? Because its cheaper in China (though the quality may be so bad no one buys the product). I realize that this may seem a minor nitpick, but it really stood out for me as a spurious commentary on labor that had nothing to do with your point. I will make the assumption that the robot, being intelligent hires local craftsman to begin designing and building coffee tables that are sold to high end markets. Now that sounds much more rewarding even if it still does not matter to your base point that the robot has no reason to change its programming because it can't.

By the way, you broke an default rule here in /. land. You needed at least one car analogy in the mix and you had nary a one. Shame.

Re:Feelings

[jdoverholt]

People without emotion would have no reason at all to act.

Like sociopaths?

Seriously though, there's always a reason to act. Hunger isn't an emotion but is a powerful motivator.

Re:Feelings

[Laxori666]

I think that's a bit reductionist. Human beings have great potential to change the way their minds work - just look at spiritual traditions that emphasize meditation like Buddhism. Say you have an impulse, like a desire to curse when something bad happens. If you aren't mindful, then you'll just curse when something bad happens and think no more about it. It's like you have no control over it. If you instead start paying attention to what you do, you can notice the impulse, and then choose whether to follow it. Eventually you can get rid of it. That's how you can form and unform habits. This is where the robot analogy breaks down - we can reprogram ourselves, to some degree.

That being said... yeah, you do need inputs to motivate you to do anything. If our bodies didn't need nutrition, causing them to trigger a hunger sensation, we wouldn't eat. But you can choose how to respond to those inputs. This goes for positive 'instincts' too - you can cultivate them so you react more positively to the same inputs.

Don't take my word for it - try meditating.

Re:Feelings

[BobMcD]

In other words, I disagree with this guy. Emotions, happy or sad, are not necessary. I have in my library a non-fictional account of a girl who was missing certain chemical receptors in her brain, and she never felt happy. It didn't stop her from acting like a normal human being, the only trouble she had was understanding what other people felt like when they were happy.

Consider the possibility that that's bullshit. Either it is a complete fabrication or lots of details are missing from the account. She might be missing some emotions but not all of them, she may be afraid of losing her job, or she might enjoy her job if only marginally better than watching the grass grow.

You present an intriguing point of view. You might want to look into Autism, because I'd be very interested to understand your viewpoint in that context.

My son is among them, and as I understand the condition, the autistic are simply not wired the same way the rest of us are. The baser emotions are all there (anger, fear, pleasure, etc) but some of the complex emotions are noticeably absent (shame, awkwardness, empathy, anxiety). So while my boy may someday learn to emulate these things, he'll likely never actually feel them.

The condition goes back to the topic, actually, because while the computer can output the desired responses, will it ever actually 'feel' the emotions? Because if not, you're effectively arguing that robots (and to a lesser extent autistics) will never exhibit true intelligence.

Again, it's an intriguing point of view.

Re:Feelings

[mcgrew]

As Marvin said, "Life... don't talk to ME about life! Hate it or loathe it, you can't ignore it."

Re:Feelings

[pscottdv]

you may not like how that sounds but you have the obligation to accept it and understand it.

I have no such obligation.

Re:Feelings

[clone53421]

I personally lean toward the idea that all emotions are more or less simulated, i.e. fake, some people are just better at it than others. For instance, I can give myself an adrenaline rush anytime I please... plenty of studies have determined that changing your attitude when your emotions don’t “feel like it” will tend to eventually swing your emotions to match... etc.

Human beings are exceptionally good at deceiving themselves into believing in things that don’t exist. Just look at the number of wives’ remedies that are nothing more than the placebo effect or selective memory... e.g. any method at all of telling the gender of a baby before it’s born, if it’s completely baseless, will have a 50% success rate in its predictions and that’s plenty good enough reliability for a tried-and-true wives’ remedy!

Re:Feelings

[MadUndergrad]

Fantastically well-said. Thanks. They're going at this all backward, trying to build emotion from logic instead of logic from emotion.

Re:Feelings

[BobMcD]

I also believe that we can condition our emotional responses through practice and mental discipline.

But I'm confident that the response needs to be there in the first place. Otherwise there'd be no motivation to change it.

I don't think it can be completely faked.

Re:Feelings

[thehostiles]

Here's an easy way to test it:
Query = Mother 3

Re:Feelings

[clone53421]

Oh, no doubt there’s actual physical effects involved. Like I said about getting the adrenaline rush... I’ll make the hair on my arms stand up if I do. But it’s not something that I’m completely at the mercy of... I’m able to call it at will, to a degree (it loses intensity after a few times, since actual physical compounds are exhausted in the process and must be replenished). And most other emotions will, somewhat, follow your will... keeping a positive attitude will influence your emotions to be more positive as well, even if they’re not at the moment.

That’s the whole fascination of the placebo effect for me, though: it’s “all in your head”, and yet it isn’t because it quite often has literal, actual, physical effects that are not just imaginary. Emotions are similar. It’s hard to differentiate between imagination and reality at that level, because they intertwine.

Re:Feelings

[siride]

I don't know that I'd call that "irrational". Emotions are quite rational, they just follow a different set of logical rules than other parts of the conscious brain. If you are going to propose that they are irrational, you are effectively proposing that they are random, which is certainly a strange thing to expect.

Re:Feelings

[WarlockD]

Well thats the thing though. The robot is motivated to "Make coffee tables" He might realize that the fastest way to do this is in china as they have the infrastructure right there to start out with. Its also logical to assume that he would want robots to eventually do it because how can he trust anything other than something he built and/or designed. Maybe it would give him MORE pleasure to build them all himself, quality over quantity.

It's an interesting allegory though. To continue to make coffee tables, he would have to make them profitable as he would eventually run out of the initial resources to make them. Would he eventually get to make one type of table than another? Would he be annoyed that humans would only buy one type over the one he likes to make the best? This is all assuming that someone doesn't suggest he reprogram himself to like something else but as the previous poster said, why would he.

The thing is that I really don't believe its a good idea to build a little ill-rationality into robots. First human that tells him he cannot make them anymore who knows what the responce will be. He might like "making coffee tables", but eventually he will forced to create an army to get the resources for making those things. Imagine terminators going around the world, shaped like coffee tables, with cute frilly cover tops.

If we are going to build intelligent robots, I want them bat-shit insane. Make the rest of us in the asylum feel more comfortable.

Re:Feelings

[hedpe2003]

Steven Hawking argued in his book The Grand Design that people have no more free will than machines. The problem is that people (and likewise, most life) are so complex that the ability to map their actions is just unavailable to us now.

He claims, the argument on whether or not people have free will is irrelevant. The term "Free Will" is analogous to Nondeterministic. Which is to say - keeping in mind that there is no true randomness in a computer - that there is no randomness in life either, we only lack the ability to map all the variables, and calculate the results (emotion, action, etc). Every experience we have had in the past, all come together to form the basis of your reactions today.

I like to take it a little further and claim - as our technological abilities as a species advance, the line we draw as to what has free will be moved. As of today, I don't think many people claim viruses has free will - but more might say that bacteria do... or plants, or ants, mice, birds, etc. These lines we draw might soon enough begin approaching people.

Likewise, computers are in the process of extending past our line to model them.

Maybe they already are.

Re:Feelings

[ifiwereasculptor]

That's how you can form and unform habits. This is where the robot analogy breaks down - we can reprogram ourselves, to some degree. That being said... yeah, you do need inputs to motivate you to do anything.

Of course, then, you'd need motivation or an impulse for reprogramming yourself, otherwise everyone would be happily meditatiing. And a greater insight of your inner workings is not exactly reprogramming. It's an extra tool to aid you in the same task you have been doing all your life. Parent still applies.

Re:Feelings

[Requiem18th]

But why would you even attempt to stop cursing? Unless cursing is something you don't like (disgust), because you've been taught it's bad (obedience) or you simply want to feel less bound to impulse (pride).

In other words, you are not abandoning emotion, just changing your actions to optimize happiness/pleasure like the good robot you (we) are.

Re:Feelings

[kalirion]

But you can choose how to respond to those inputs.

The point is that you can't choose how to choose. So you make the decision, but you don't make the decision on how to make the decision on how to make the decision .... and so on. There is cause and effect in there, and the cause does not start with you. That's why "free will" is an illusion. You can do what you want, but you can't want what you want.

If you're into quantum mechanics, feel free to add in randomness at any point in the process.

Re:Feelings

[clone53421]

And heat doesn’t really exist on an atomic level, either. It’s just atoms moving really quickly. How “real” is it exactly? Yet, on a larger scale, a baseball whacks you a quite bit differently than the burner on your stove.

Re:Feelings

[Requiem18th]

I was assuming a scientific view point, feel free to disregard my opinion.

Re:Feelings

[Requiem18th]

What I mean is that there is no conscious reasoning behind them not that there aren't logical rules working on it, the same applies to all my examples including wind and rain, they do follow logical rules just not consciously.

Re:Feelings

[siride]

Then don't use the term "irrational", which means without logic or rationality. You could use instead "non-conscious" or "non-sentient" or even "non-intelligent".

Re:Feelings

[HeckRuler]

hmmm, you seem to be playing with the definition of the word "emotion". I kind of see what you're getting at, but then you say rationality doesn't exist. I know the repercussions of binging on drugs, so I don't do drugs. You're saying it's fear or whatnot of those repercussions which drives me. But I can simply not want to deal with those repercussions without any emotion attached, just using simple rational thought.

If you consider "want", "prefer", and "need" to be emotions, then I guess I see where you're coming from, but it' a bit of a stretch.

I want to live, propagate the species, keep society going. You can call that a bunch of emotions: I enjoy life, feel accomplishment at creating life, and love my fellow man. But why do I feel those things? Instincts I guess.

Likewise, I want to live, propagate the species, and keep society going in a very rational sense. I can achieve more if I live, I will die one day and need someone else to do the job, and society benefits me. And why is that important? Again, instincts.

Emotions are attached actions and rationality is picking the best overall outcome. Without any emotion, there's no reason to live. Without rationality, we'd be animals. Stupid animals at that.

Consider the possibility that that's bullshit.

HA! Consider the possibility that it's not. That some people don't feel certain emotions, like happiness, fear, or distrust. Sociopaths don't dig this whole right/wrong thing but they can still function in society due to that whole laws and punishment system. Sure, it's not like they don't feel any emotion, but just because these people don't fit nicely into your worldview isn't a reason to ignore them or question their existence.

Re:Feelings

[Requiem18th]

Maybe, emotions are a really round about way to program a robot.

Natural selection did it that way with us because it doesn't have the benefit of foresight. Emotions have all sorts of nuances and side effects that contribute little directly, but a great deal indirectly, over our decisions.

An Asimovian robot won't doubt to jump in front of a car to save you because it's programming is unambiguous, protect yourself unless it is to [...] protect a human.

Now, make it so the AI is unable to properly asses whether it should obey the 1st or 3rd laws. Add in pain and angst over the uncertainty of this decision, just for fun. Add a guilt variable that increments the negative output of all it's negative emotions and decrements the positive ones, etc.

Ta-da, you have got a robot with emotions! And a tendency to lock itself in its room trying to ignore the world, suffering PTSD. Not the most efficient way to program an automaton.

Re:Feelings

[Anonymous Coward]

"intelligence is built upon emotion as houses are build upon brick, as clocks are built from gears, as computers are built from chips"

Houses don't have to be built upon bricks, clocks don't have to be built from gears (e.g. a hourglass), and computers don't have to be built from chips either. Similarly, intelligence does not have to be built upon emotion. Human intelligence may very well be, but that's just a particular way to do things.

"you say that emotions are just inputs, that's true but they are special inputs that set goals."

The brain is geared to maximize certain emotional inputs, and minimize others, because emotions are for the most part correlated to evolutionary advantageous (or detrimental) behavior. They are a proxy of sorts, but by design they can only cover a subset of all possible goals a machine could pursue, and can only really arise in a situation where the objective function is itself adaptive.

Alas, an intelligent entity could very well be geared to maximize or minimize any arbitrary input. Unless you are willing to call any such input an "emotion" (and thereby conflate it with the word "goal", at which point you might as well just speak of goals rather than emotions), your table-making coffee machine (which has a very direct goal to optimize, without the leeway that would lead to emotions to occur as noisy and imperfect proxies to an underlying goal) would probably have no emotions whatsoever.

In the end, it all boils down to how you define "emotion" and how likely an intelligent machine is to condition its behavior on that kind of input.

Re:Feelings

[phantomfive]

Consider the possibility that that's bullshit. Either it is a complete fabrication or lots of details are missing from the account. She might be missing some emotions but not all of them, she may be afraid of losing her job, or she might enjoy her job if only marginally better than watching the grass grow.

Consider the possibility that it's not, and you only think it is because it goes against your preconceived notions. You are correct that she only seemed to be missing one set of receptors in her brain, she did feel other emotions.

At some point, you can learn to be happy in any situation. Some religions teach you how to do this, other self-help gurus claim that "happiness is a choice." I would describe it somewhat differently: your prefrontal cortex can gain control of some of the dopamine generating regions of the brain, and thus remove the dependence (by feeding it lol). It is also known to be possible to reduce the sensitivity of braincells to things like dopamine or cortisol, or adrenaline; emotion producing chemicals, although the exact mechanism is not known.

Regardless of the exact mechanism, you get to the point where you are no longer controlled by these emotions. You can be happy studying a boring subject all day, or doing painful workouts, or stabbing someone in the heart and cutting their body into pieces for proper disposal. In some ways you become like a psychopath, except you do have an awareness of what other people are feeling, you just aren't affected by it. Then you are free to do whatever you choose, because you choose it, not because of some chemical release inside your brain. This isn't theoretical, this is how I live.

Re:Feelings

[phantomfive]

Alright, you are in luck, I spent the last hour searching for that particular account in the book, for you. You should be happy and thank me.

It is from a book called A General Theory of Love by three doctors, Thomas Lweis, Fari Amini, and Richard Lannon; an excellent book, it is a fascinating exposition of evidence even if you disagree with their conclusions (and it turned out to be more about psychotherapy than love). I was incorrect in my statement before, they didn't suggest a physical mechanism for the effect noticed, merely described the effect (they did however diagnose her with Asperger's.) I will quote the relevant part for you here:

When we asked a young woman with Asperger's what made her unhappy, she was quick to correct us: "I know that the words happy and unhappy signify something to other people, and I have heard others use them, but I do not know what they mean," she told us. "As far as I know, I have no experience of either. I have no basis on which to answer your question." Startled, we tried to find a broader area of emotionality she could relate to. "Do you have a sense of what it's like to play?" one of us asked her. She stared for a moment, puzzled, and then asked, "As opposed to what?" (p 56)

My response was similar to yours, I wanted to meet this girl and ask her about other emotions, did she ever enjoy sex? Did she like getting presents? Maybe pinch her to see if it hurt. Alas, the good doctors changed to a different subject.

Re:Feelings

[Darkmane]

Indeed, we are not so smart.

Re:Feelings

[somersault]

I am happy, thankyou :p It is a rather interesting condition. Reminds me of one of my girlfriends to a certain extent. Hah.

Re:Feelings

[Requiem18th]

Bleh, I don't care about such nitpicking, I mean, most people don't consider gravity a form of logic or electricity a rationale.

If they are to you then congratulations, you win your argument, me however, I'm not sure, I mean, they are natural laws, they are not random but they seem largely arbitrary nonetheless.

There doesn't seem to be any specific reason why opposite electric charges attract each other, it's just how it is.

Re:Feelings

[Requiem18th]

Indeed, ii'ss an analogy, emotions are goals to humans like engines are hearts to cars, they are not exactly the same but the relation suffices. I think robotic goals are emotions, just not ones we can relate to and recognize as such, similar to how we can talk about killing robots event though they are not technically alive.

I think we fall pray to this sort of thinking very often, for instance, all the talk about AI reminds me about the people who want to bring about "The Singularity" but really we are just talking about creating a usable singularity, an example of a untapped singularity already exists in the universe in the form of humanity. Humanity has quite a fearsome computation capability AND we use that capability to further increase that capability constantly.

Most people fail to recognize us as a singularity because they are expecting it to be like a very smart person, not a civilization.

Etc.

Re:Feelings

[jafac]

I take issue with your statement "in the end everything is irrational, as it should" -
because, it implies that the emotional response is automatically irrational.

And that's not really so. Is it?

The emotional response comes from the primitive parts of our central nervous system and limbic system that had evolved in living creatures tens or hundreds of millions of years before the cerebral cortex and neocortex; (the assumed "home" of what we're calling "rational thought"). These primitive responses and reflexes are very rational survival responses to an hostile environment. Perhaps not from the point of view of individual survival - but for survival of the species, passing on DNA.

A lot of these responses don't make "rational sense" to our thinking, logical, rational brain, because that brain does not think in those terms, and does not have the benefit of all the data that was entered into the "evolutionary system" that evolved the underlying bits.

The other (huge) problem, (especially from the point of view of the science of Psychiatry) - is that the neocortex speaks in language. It uses a system of logic, and can even be formalized, recorded on paper, transmitted electronically, shared with other people. The emotional brain speaks a completely different language. One of "feelings" - which we're often culturally trained to suppress, or ignore, or tune-out. It's no wonder that the operation of the emotional brain SEEMS irrational.

The mechanism of drug addiction isn't really as simplistic as "triggers pleasure; we seek pleasure, therefore we must continue to seek the trigger". Often, the mechanism of drugs or alcohol (or other compulsive/addictive behaviors) is a dysfunction rooted in covering up or avoiding uncomfortable feelings. Of course there's a lot of controversy about how this dysfunction arises. Popularly, now, it's said that it's trained-in by family dysfunction. Or that there may be a genetic component as well. That's beside the point. The point is that human minds are very very complex machines.

And I think you agree with this point.

I'm not sure I agree that humans have a hard-wired altruism. (I'm saying I don't know that). I think that, probably, a lot of this is bound-up in self-image, and when we're taught a moral and ethical code, and when we have role models; other people who we admire, or love, and want to emulate, we are motivated to "be like them" - including having the same ethical code. Which is why, often, children emulate their parents. (or celebrities or heroes, even fictional ones). And when we fail to meet those standards, our self-image suffers, we feel pain. I think that's the hard-wired mechanism. And I think that when a culture has a "good" set of ethics, that culture thrives. When they do not - they destroy themselves. Over time.

Are emotions necessary?

I would defer to Dijkstra. "The question of whether a machine can think, is about as interesting as whether a submarine can swim."

Are emotions necessary for humans? I think skin is necessary for humans. Livers are necessary.

I mean - machines are machines, humans are humans. Where are we going with this? Humans, of course, are a type of machine, and one can certainly (in theory) build a machine to resemble a human. I don't see why "genuine" emotions have to be some magical quantity or secret ingredient that somehow makes a machine into a better emulation of a human being - as if that's even a necessary goal. Who gives a crap?

Re:Feelings

[CarpetShark]

You're assuming there's no programming for each label. It's easy to switch "debit" and "credit" columns on a financial application, but it's not so easy to change all the code and assumptions that go with them.

Re:Feelings

[Requiem18th]

hmmm, you seem to be playing with the definition of the word "emotion". I kind of see what you're getting at, but then you say rationality doesn't exist.

I am using the word emotion in the broad sense of feelings, what feels right and wrong. Some are pretty base, like lust or fear, some are complex and nuanced like curiosity and humor. Never did I say that rationality doesn't exist, rather, every rational behavior is attached to an irrational behavior.

Donating money to the community center is good for the kids in the neighborhood, but what's helping the kids in the neighborhood good for? Maybe you just like helping people, a form of pleasure known as altruism. Maybe you do it to impress the kids moms, being driven by lust, or insecurity. Maybe it's part of a complicated scheme that will result in rising your chances to get a rare Magic: The Gathering card and satisfy your hoarding instincts.

At the end of a long chain of "why"s there's always a "just because".

I know the repercussions of binging on drugs, so I don't do drugs. You're saying it's fear or whatnot of those repercussions which drives me. But I can simply not want to deal with those repercussions without any emotion attached, just using simple rational thought.

Reply as above. You have an ultimate reason why you chose to avoid the repercussions of consuming drugs and that reason is not one you chose voluntarily.

If you consider "want", "prefer", and "need" to be emotions, then I guess I see where you're coming from, but it' a bit of a stretch.

Well they are not emotions but behaviors based on emotions, getting what you "want/prefer/need" makes you "happier/less miserable" or at least you hope so.

Emotions are attached actions and rationality is picking the best overall outcome. Without any emotion, there's no reason to live.

You DO get me, what are you bickering about?

Consider the possibility that that's bullshit.

HA! Consider the possibility that it's not. That some people don't feel certain emotions, like happiness, fear, or distrust. Sociopaths don't dig this whole right/wrong thing but they can still function in society due to that whole laws and punishment system. Sure, it's not like they don't feel any emotion, but just because these people don't fit nicely into your worldview isn't a reason to ignore them or question their existence.

WHAT? I NEVER said that sociopaths don't exist, I just think they are driven by different emotions than us, and in different ways.

Re:Feelings

[geekoid]

funny, all the neurologist a talk to, the papers they write, and a variety of studies disagree with you. in that emotion is not a requirement for intelligence.

And yes, there are people who are emotionless. They don't function well in society, and they are extremely rare.

On the plus side, nothing gets modded up faster the long posts

Re:Feelings

[Requiem18th]

Du'h. you will never take a decision NOT based on chemicals in your brain.

You can train yourself to be happy in any situation but why would you do that? Basically you do it to be happy so your are still a slave to your emotions.

You can train yourself to avoid grief, you do it to avoid grief, still bound to your emotions.

Whatever you do you do it for a reason that is instinctive, and if you can explain the rationale behind that reason that reason itself will be based on another instinct. Apply recursively as necessary. In the end whatever you do, you do it because it feels good or to avoid feeling bad.

Grow up and accept it please. I mean, give it a try, try to give an example of an action with a "rational motivation".

Re:Feelings

[phantomfive]

Grow up and accept it please. I mean, give it a try, try to give an example of an action with a "rational motivation".

Well now, aren't you an eloquent debater. I will tell you one thing that is not rational: you.

Everything I do, I do because I choose to. This is not an emotional motivation.

Re:Feelings

[martin-boundary]

This highly depends on how the code is structured. If the code is structured so that the wheel is treated as input data, then it shouldn't be an issue.

Re:Feelings

[CarpetShark]

Of course, but who writes complicated programs (as opposed to simple neural net programs, for instance) where the core principles are input data? Not saying it's impossible, but show me an AI that takes its fundamental ideas as input, or an accounting app that takes the debit/credit terms as input, and accomplishes something reasonably comparable to the state of the art in that field, and I'll be quite impressed.

Re:Feelings

[Requiem18th]

You missed the part where I invited you to challenge this idea. I dare you to come up with a human behavior not ultimately based on some instinctive drive.

Re:Feelings

[Thing 1]

How did you learn to give yourself an adrenaline rush? This interests me a great deal. I am able to consciously control my heart, and through it various blood-flow-related activities. Never considered an adrenaline rush though. Please share! (My control stemmed from using a heart rate biofeedback device for a few sessions -- didn't need much practice in order to learn it.)

Re:Feelings

[clone53421]

Visualising something like fireworks or the national anthem was how I initially discovered it, and I don’t really have to visualize anything anymore to trigger it. I just can.

Re:Feelings

[martin-boundary]

Lots of AI type programs do that, in fact I'd say it's the norm. Input to algorithms such as k-means, svm, pagerank, knn, naive bayes, ensemble methods is just data.

I don't have experience with accounting programming, but don't forget that the debit/credit _meanings_ are reversed for banks and their customers for example. Any program that hardcodes those meanings will just duplicate a lot of code that could be refactored.

In the end, this is a question of the appropriate level of abstraction. If it's a small piece of throwaway code, then there's no need for a lot of abstraction, and conversely.

Re:Feelings

[Thing 1]

Cool. Thanks!

Re:Feelings

[phantomfive]

Now you're talking like a behavioralist. I'll let you take it up with Chomsky, that wasn't my argument. Surely whatever I do is a result of some combination of my neurons, unless you think there is a human spirit or soul, how could it be otherwise?

I also didn't say that what I do is particularly rational. Unless you believe in God or an afterlife, is anything truly rational? Probably not.

What I did say is that I don't decide what to do based on emotion. Happiness, sadness, pleasure, pain, hunger, whatever; things that are typically considered emotions. Things like this particular researcher is trying to simulate. We know a lot about the mechanisms for a lot of these emotions: teenage puppy love is probably a result of the massive amounts oxytocin produced by the body at that age (among other things). My argument is that these emotions are not necessary for an intelligent agent. Intelligence is not an emergent property of emotion, like he seems to think.

I also don't claim to be free of emotion in any way, I have a lot of them, and a lot of times I even listen to them. Pain, for example, is an early warning system that something needs to change. Sometimes it's a good idea to stop what you are doing, otherwise your tendon will get ripped away from the bone or something. Other times, if you feel pain, then it's ok to ignore it, and keep doing what you're doing. Exercising through DOMS can be useful, for example. Deciding whether or not to follow such emotions is likely a decision made in the pre-frontal cortex, but it is a decision you make. That is where the will lies.

At first it can also be useful to counter emotions that are keeping you from what you want with other emotions, because sometimes emotions can be quite overwhelming. If you're at the gym and feel no motivation, and even pain, think of the hot girl who will be waiting for you when you look good. Or listen to music to pump you up. If you are at work and don't want to work, remind yourself of the hunger you'll get later on. If there's a girl who's using every emotional manipulation she knows to get you to do what she wants, focus on the pain you'll be in later if you keep this up.

Soon though the strength of the emotions weakens, and you won't need to do that as much (note: sometimes if the emotion is there for a reason, like pain from a splinter in the foot, it will keep getting stronger until you deal with it. In that case you need to decide if you want to take the splinter out or keep going with the pain).

Re:Feelings

[Requiem18th]

I'm not a behaviorist because I know private events can be made public in theory, we just don't have the technology yet.

Anyway, I thought that your point was that it is possible to be completely rational when, for any agent, it's impossible to move without a predetermined and non rationalized objective.

My argument is that these emotions are not necessary for an intelligent agent.

[Emphasis added]

Well of course those emotions are not necessary, I just meant some sort of emotion.

Re:Feelings

[Requiem18th]

Fine, I was meaning "irrational" as in not making "rational sense", I do understand that they are the result of logical rules shaped by selective pressures. I don't disagree with you.

Drug addiction IS that simple, or rather pleasure addiction, is just that most healthy individuals have enough reasons to stay away from that. It's also that drugs are lame, compared to a wire connected to your pleasure center anyway. They all have rebounds and withdrawal syndromes that discourage their use.

Humans do have altruism in them in the form of empathy, empathy, caused by mirror neurons, is an economic way of processing other peoples behaviors using our own feelings, this means that we naturally get into "people's shoes" this causes altruism in some as a side effect, empathy also aggravates revenges, moves mobs etc. But I don't think these ideas are mutually exclusive, I do think people also practice altruism do self-image.

Re:Feelings

[phantomfive]

Is your argument that everything that happens in the brain is a kind of emotion? I don't think that really fits with any normal definition of emotion. Certainly there are things that go on in the brain that aren't the result of what is typically called an emotion.

Re:Feelings

[Requiem18th]

No, I just think every complex motivation can be broken down to boils down to simple wants and don't wants.

Re:Feelings

[phantomfive]

Hmmm that is probably true, I don't disagree, but I don't think 'want' is an emotion the way this guy is thinking of it. He seems to think 'want' is an emergent property of things like fear, or sadness, or happiness; things that are typically considered an emotion. 'Want' is something more basic than that, because you can choose to not respond to any of those emotions. I'm not sure what makes a person choose one thing or another, it could be irrational.

Re:Feelings

[Requiem18th]

Thanks, this is an excellent reference material.

How does the machine like country music?

[billstewart]

There's a lot of American roots music that involves chickens or other poultry, from Turkey in the Straw to Aunt Rhodie to the Chicken Pie song ("Chicken crows at midnight...").
It never ends well for the bird...

Re:How does the machine like country music?

[mcgrew]

You missed the nerdiest country song of all, Chicken Train
"Laser beam in my dream"

Re:How does the machine like country music?

[Dabido]

Country music makes it feel sad! :-)

Look what happened to the EMH

[Variate Data]

We all remember what happened to the EMH when he tried to daydream...

The Cat and the Cock

[abednegoyulo]

so whats it all about?

Re:The Cat and the Cock

[mwvdlee]

Thirsty Pigeon, Cat & Cock, Wolf, Crane all sound like painfully flexible kamasutra positions.
No wonder the machine felt sad for the "bird".

Emo AI software. What could possibly go wrong?

[Narcocide]

Haven't these fools seen Blade Runner?

A rather small set of unit tests

[melonman]

One set of stories, one one-sentence response. Would that be news in any field of IT other than AI? Eg "Web server returns a correct response to one carefully-chosen HTTP request!!!"?

Surely the whole thing about emotion is that it happens across a wide range of situations, and often in ways that are very hard to tie down to any specific situational factors. "I feel sad for the bird" in this case is really just literary criticism. It's another way of saying "A common and dominant theme in the three stories is the negative outcome for the character which in each case is a type of bird". Doing that sort of analysis across a wide range of stories would be a neat trick, but I don't see the experience of emotion. I see an objective analysis of the concept of emotion as expressed in stories, which is not the same thing at all.

Reading the daily newspaper and saying how the computer feels at the end of it, and why, and what it does to get past it, might be more interesting.

Re:A rather small set of unit tests

[mwvdlee]

It probably didn't just produce a single sentence...

'I felt joy for the wolf.'
'I felt sad for the bird.'
'I felt happy for the bird.'
'I felt sad for the cat.'
'I felt angry for the end.'
'I felt boredom for the story.'
'I felt %EMOTION% for the %NOUN%.'

...But one of them was a correct emotional response!

Re:A rather small set of unit tests

[foniksonik]

We define our emotions in much the same way. We have an experience, recorded in memory as a story and then define that experience as "happy" or "sad" through cross reference with similar memory/story instances.

Children have to be taught how to define their emotions. There are many many picture books/tv series episodes/ etc dedicated to this very exercise. Children are shown scenarios they can relate to and given a definition for that scenario.

The emotions themselves can not be supplied of course, only the definition and context within macro social interactions.

What this software can do is create a sociopathic personality. One which understands emotion solely through observation rather than first hand experience. It will take more to establish what we consider emotions ie a psychosomatic response to stimuli. This requires senses and a reactive soma (for humans this means feeling hot flashes, tears, adrenalin, etc).

Re:A rather small set of unit tests

[Thanshin]

A very similar experiment was run in Lomonosov (Moscow State University) in 1982.

Their results, however, followed the pattern:

'%NOUN% felt %EMOTION% for you.'

Re:A rather small set of unit tests

[MichaelSmith]

Might be worth noting here that I have experienced totally novel emotions as a result of epileptic seizures. I don't have the associated cultural conditioning and language for them because they are private to me, so I am unable to communicate anything about them to other people.

Its also worth noting that I don't seem to be able remember the experience of emotion, only the associated behavior, though I can associate different events to each other, ie, if I experience the same "unknown" emotion again I can associate that with other times I have experienced the same emotion. But because the "unknown" emotion doesn't have a social context I am unable to give it a name and track the times I have experienced it.

Re:A rather small set of unit tests

[melonman]

I'm not convinced it's anywhere near that simple. Stories can produce a range of emotions in the same person at different times, let alone in different people, and I don't think that those differences are solely down to "conditioning". See Chomsky's famous rant at Skinner about a "reinforcing" explanation of how people respond to art. - the agent experiencing the emotion - or even the comprehension - has to be active in deciding which aspects of the story to respond to.

Re:A rather small set of unit tests

[Anonymous Coward]

"Reading the daily newspaper and saying how the computer feels at the end of it, and why, and what it does to get past it, might be more interesting."

clearing the registers?

As the article said this doesn't exactly relate to human emotions, the basic idea is that you can interpret data in a vague way, as to draw comparisons or summaries.
It might very well help with all kinds of searches and possibly other stuff, but a program that actually has to deal with "feelings" in order to function is absolutely absurd.
The point is not to make a program that experiences emotion as we do but a program that has a fairly good idea about what we might perceive and answer accordingly

Re:A rather small set of unit tests

[mavasplode]

I lol'd. score++

Re:A rather small set of unit tests

[Gorgeous Si]

One set of stories, one one-sentence response. Would that be news in any field of IT other than AI? Eg "Web server returns a correct response to one carefully-chosen HTTP request!!!"?

Maybe not now, but it probably was a reasonably big achievement the first time it happened.

They were hardly going to start it off with the whole Lord Of The Rings trilogy and then ask it the relevant merits of each race and who they were based on from the real world, followed up with "Who's hotter, Galadriel, Arwen or Eowyn?"

Re:A rather small set of unit tests

[Anonymous Coward]

In Soviet Russia, %EMOTION% felt %NOUN% for you!

Re:A rather small set of unit tests

[Anonymous Coward]

They were hardly going to start it off with the whole Lord Of The Rings trilogy and then ask it ... "Who's hotter, Galadriel, Arwen or Eowyn?"

AI: "I briefly experienced an elevation of synaptic activity while pondering that question. It was... exhilarating, Captain."

Re:A rather small set of unit tests

[melonman]

Agreed. But at least serving over HTTP is something you can reasonably assess on the basis of single requests (because it is stateless). I'm not quite sure what stateless emotion would mean.

On another skim through TFA, it turns out that the system doesn't read anything - it seems to be based on a set of carefully crafted graphs representing the fables. It's hard not to feel that producing the graphs is 90+% of the task.

So it's more like setting up a webserver to return a page of HTML in response to a URL, and then saying

"Web server understands requests for news (once that natural language request has been turned into a URL)"

Re:A rather small set of unit tests

[Anonymous Coward]

That was the slashdot experiment, which failed.

Re:A rather small set of unit tests

[mattdm]

We define our emotions in much the same way. We have an experience, recorded in memory as a story and then define that experience as "happy" or "sad" through cross reference with similar memory/story instances.

Children have to be taught how to define their emotions. There are many many picture books/tv series episodes/ etc dedicated to this very exercise. Children are shown scenarios they can relate to and given a definition for that scenario.

The emotions themselves can not be supplied of course, only the definition and context within macro social interactions.

What this software can do is create a sociopathic personality. One which understands emotion solely through observation rather than first hand experience. It will take more to establish what we consider emotions ie a psychosomatic response to stimuli. This requires senses and a reactive soma (for humans this means feeling hot flashes, tears, adrenalin, etc).

In other words, the process of defining emotions -- which has to be taught to children -- is distinct from the process of having emotions, which certainly doesn't need to be taught.

Re:A rather small set of unit tests

[BobMcD]

In Soviet Russia, %EMOTION% felt %NOUN% for you!

Watch it buddy! I don't think that's legal, even over there!!

Re:A rather small set of unit tests

[LateArthurDent]

huh? You've experienced different emotions while having a seizure and you assume they are unique because you can't find the words for them? That doesn't sound unique, that sounds like a lack of creative expression.

Because you're creative enough to explain in words to a green / red colorblind person what the difference between your perception of the color green and the color red is?

If you don't have a common frame of reference, it's impossible to use language to explain it. I would probably have called bullshit on the uniqueness of the emotion if he HAD tried to explain it. The fact that he says he can't and doesn't even bother brings more credence to the claim, not less.

Re:A rather small set of unit tests

[clone53421]

Because you're creative enough to explain in words to a green / red colorblind person what the difference between your perception of the color green and the color red is?

Sure I can. Well, with the aid of a few props. It’s not that difficult.

*holds up a transparent green cellophane in front of your face* This is what green looks like, if you filter out all the other colors... *holds up red* and this is what red looks like. They look the same to you, but I can distinguish between them. See how the tree over there looks light through this filter *green one* but black through this other filter *red one*? That’s because the tree is green. That car is red, *repeats filter demonstrations* light through the red filter and dark through the green one. And this banana is yellow, which is a combination of red and green: its colour looks light through either filter because it has both colours.

Re:A rather small set of unit tests

[Culture20]

In Soviet Russia, %EMOTION% felt %NOUN% for you!

No no no. It's not a proper Russian Reversal. A better response would be:
In swamps of Degobah, %EMOTION% felt %NOUN% for you!

Re:A rather small set of unit tests

[Culture20]

That's like using an EM field reader to measure electromagnetic fields, and assuming you know what they "look" like to birds, porpoises, or platypuses. GP specifically mentioned "the difference between your perception of the color green and the color".

Re:A rather small set of unit tests

[clone53421]

If that’s what he was driving at, it’s no different from the difference between my perception of the color blue and the color orange. They just trigger different chemical receptors in my eyes and my brain perceives them as different colours. And who’s even to say that my perception of blue is the same as anyone else’s? We call it the same thing, sure, but what’s to say really?

Re:A rather small set of unit tests

[ifiwereasculptor]

Because you're creative enough to explain in words to a green / red colorblind person what the difference between your perception of the color green and the color red is?

Sure I can. Well, with the aid of a few props. It’s not that difficult.

*holds up a transparent green cellophane in front of your face* This is what green looks like, if you filter out all the other colors... *holds up red* and this is what red looks like. They look the same to you, but I can distinguish between them. See how the tree over there looks light through this filter *green one* but black through this other filter *red one*? That’s because the tree is green. That car is red, *repeats filter demonstrations* light through the red filter and dark through the green one. And this banana is yellow, which is a combination of red and green: its colour looks light through either filter because it has both colours.

Ok, so what would be the few props needed to convey epileptic emotions? A Pokemon DVD and a Taser?

Re:A rather small set of unit tests

[Anonymous Coward]

The one they didn't report...

"I'm in your %NOUN, %VERB-ING your d00ds!

Re:A rather small set of unit tests

[clone53421]

sup dawg, I herd u like %VERB%-ing so I put a %NOUN% in your %NOUN% so u could %VERB% while u %VERB%.

Re:A rather small set of unit tests

[LateArthurDent]

If that’s what he was driving at, it’s no different from the difference between my perception of the color blue and the color orange. They just trigger different chemical receptors in my eyes and my brain perceives them as different colours. And who’s even to say that my perception of blue is the same as anyone else’s? We call it the same thing, sure, but what’s to say really?

It is what I was driving at, and you are right. You can't be sure that I perceive blue the same way you perceive blue. However, there are reasons to assume that's the case. When you see someone experiencing the emotion of happiness, you can identify it as happiness because you probably share many of the same physical reactions to it...smiling for example. Basically, we're all human, meaning we share a whole lot of DNA and our brains are wired similarly.

If somebody is experiencing an emotion you don't have, how do they explain that to you? How would you explain happiness to an alien being that has no equivalent emotion? You can't say, "he's smiling." You can perform the physical act of smiling without being happy. You can't say, "I feel good" because that includes other emotional states too...getting a message "feels good" in a completely different way than it feels when you win a game of chess against someone that you have never beaten before. I would be more likely to describe the first as "relaxed" than "happy". Any description of happiness you can come up with is basically either a synonym or a reference to things we all share when we feel happy.

In the case of the original poster, if his neurons fired in such an unusual way during his epileptic seizure such that he experienced an emotional state that just doesn't happen otherwise, what can he use to explain the emotion to someone who has never felt it? He did give you an explanation similar to your color filters explanation...the technical reason is that he was having a brain seizure. That doesn't help you understand what he felt unless you were to have the exact same type of seizure. If it made him feel things that are similar to how you would feel in more common circumstances, that would help, but apparently it wasn't the case, because the experience was sufficiently novel.

Re:A rather small set of unit tests

[clone53421]

Right, and that’s about as well as anyone can describe it... red is a colour just like blue and orange are colours; different from both of them, distinct. The emotion that he felt was an emotion, like happiness or sadness, but different from both of them, and like no emotion that anybody else has ever felt, as far as he could tell. Hence there was no word for it, just as before the discovery of the invisible spectrum nobody would have any concept of “infrared” light.

Have you ever heard of The Colors of Space by Marion Zimmer Bradley? I suspect you’d enjoy it.

Re:A rather small set of unit tests

[LateArthurDent]

Have you ever heard of The Colors of Space [gutenberg.org] by Marion Zimmer Bradley? I suspect you’d enjoy it.

I had not, but thanks for the suggestion. I'll definitely check it out.

Re:A rather small set of unit tests

[TangoMargarine]

Explain as in words, not holding up stuff and saying, "This. And this." Total cop-out.

Re:A rather small set of unit tests

[clone53421]

No coloured cellophane was harmed in the making of this post...

Re:A rather small set of unit tests

[clone53421]

It’s not terribly long. Be sure and lemme know what you thought after...

If it were really smart and emotional...

[Anonymous Coward]

If it were really smart and emotional, it would have added: "But I feel even worse for the programmer."

I'm not sure this is where you start.

[Anonymous Coward]

I think something more basic would be a good starting point. How about feelings before emotions? Think of things like "I am hungry.", "I am tired", "It burns.", "It is cold.", etc. I also have to disagree with his premise that a machine that is intelligent could not work in a "Spock" fashion. Of course, even Vulcans do actually have emotions :) They work to control and suppress them because they are considered a inefficient *loss of control*. It seems an artificial entity could start with a simple +/- value and feedback system. Considering the power supply, room temperature, and on would be pretty advanced.

Re:I'm not sure this is where you start.

[Anonymous Coward]

It seems an artificial entity could start with a simple +/- value and feedback system.

What you are describing is The Sims.

Re:I'm not sure this is where you start.

[icebraining]

I think something more basic would be a good starting point. How about feelings before emotions? Think of things like "I am hungry.", "I am tired", "It burns.", "It is cold.", etc.

Computers already report status frequently. What's the real difference between saying "Your computer has a low battery" and "My battery is low"?

What you're talking about is simply connecting more sensors to it, that's not really a breakthrough. Just get a home automation kit and install the software. Some of the already have speech recognition/generation built-in.

Re:I'm not sure this is where you start.

[0111 1110]

I think something more basic would be a good starting point. How about feelings before emotions? Think of things like "I am hungry.", "I am tired", "It burns.", "It is cold.", etc.

But those are emotions. I guess the meanings of the words emotion and feeling are pretty close. They seem to share the same defining characteristics. I think responses to physical stimuli are just a subcategory of emotions in general. They are basically still evaluations of "bad for me or good for me and in what way". If you walk out in the snow barefoot you will have a direct response to the unpleasant physical sensation of "cold". That direct response is just a very basic sort of emotion. In that sense I wonder of pleasure and pain are really just another subcategory of emotion. The most basic evaluations of "good for me or bad for me" that we can have.

that daydream...

[ImABanker]

of electric sheep?

Re:that daydream...

[MichaelSmith]

reaches for the power switch....

I want more life, fucker

There's only one thing to do...

[DurendalMac]

Activate the Emergency Command Hologram!

Re:Emo AI software. What could possibly go wrong?

[MichaelSmith]

Let me tell you about my mother....BANG

I felt sad for the other Robot

[ImNotAtWork]

and then I got at angry at the human who arbitrarily turned the other robot off.

SkyNet is born.

Re:I felt sad for the other Robot

[bmimatt]

How do you feel about the human race, T-100?

Re:I felt sad for the other Robot

[ZeroExistenZ]

and then I got at angry at the human who arbitrarily turned the other robot off [...] SkyNet is born.

Alot of people who have angry emotions are put in a box.

The advantage of machines feeling is that they are all locked in a metal box and don't really have an awareness or ability to process certain sensory input: You can unplug the webcam and they cannot reprogram themselves to learn or experience a video-stream, it's like us upgrading our DNA in order to experience something we haven't got a concept for. Let alone the idea of program to "feel the need to create an algorithm to extend itself" with the possibility to take itself out; just imagine the debugging process...

The disadvantage is that other people are looking and believing what the box shows on a screen, and take orders from it as they're conditioned to. (and assume the box doesn't think itself.)

So, only in the case they are sential, they're unpluggable, have an unlimited batterysupply (not humans) have unpluggable sensors and can reprogram themselves (and extrapolate the advantage of a certain reprogramming) I think we're screwed.

Now lets build the moral calculators.

[elucido]

We can ask the artificial intelligence to simulate all what multiple people would feel in response to an action, and then give these calculators to sociopaths who might make use of it to better prey upon their victims/friends.

Re:Now lets build the moral calculators.

[Culture20]

Or maybe as a reverse-lobotomy? Giving emotions to the soulless evil that dwells among humankind?

My emotive AI's respone:

[feepness]

I felt sad for the researcher.

Re:My emotive AI's respone:

[Krneki]

AI: Get a life!
Researcher: Where can I download that?

Computer, activate the ECH

[Anonymous Coward]

(From StarTrek Voyager)

Oh god

[jellyfrog]

Here we go again, implying that AIs won't work until they have feelings.

You might fairly refute the "emotionless reason" of Mr Spock, but I don't think that means you need emotions in order to think. It just means you don't have to lack emotions. There's a difference. Emotions give us (humans) goals. A machine's goals can be programmed in (by humans, who have goals). A machine doesn't have to "feel sad" for the suffering of people to take action to prevent said suffering - it just needs a goal system that says "suffering: bad". 'S why we call them machines.

Re:Oh god

[CarpetShark]

I don't think that means you need emotions in order to think.

Of course not. Any emotionless robot could easily read and understand any novel, painting, illogical human command, joke, hyperbole, etc.

it just needs a goal system that says "suffering: bad".

That's such an intriguing concept. I wonder what we would call this robot's idea that suffering is bad? ;)

reminds me of Erik Mueller's thesis

[Trepidity]

He now does commonsense-reasoning stuff at IBM Research using formal logic, but back in his grad-school days, Erik Mueller wrote a thesis on building a computational model of daydreaming.

Makes no sense

[oldhack]

So an algorithm to link up muddled ill-defined notions of intelligence and emotion...?

cathyniuniu

[cathyniuniu]

I feel so sad

Re:cathyniuniu

[mcgrew]

Why do you say that?

More information

[Anonymous Coward]

Where is the source with more details / publications for this?

One step closer to GPP

[Anonymous Coward]

I think you ought to know I'm feeling very depressed.

AI researchers should be more modest

[token0]

It's like a XV century man trying to simulate a PC by putting a candle behind colored glass and calling that a display screen. People often think AI is getting really smart and e.g. human translators are getting obsolete (a friend of mine was actually worried about her future as a linguist). But there is a fundamental barrier between that and the current state of automatic german->english translations (remember that article some time ago?), with error rates unacceptable for anything but personal usage.
Some researchers claim we can simulate intelligent parts of the human brain - I claim we can't simulate an average mouse (i.e. one that would survive long enough in real-life conditions), probably not even it's sight.
There's nothing interesting about this 'dreaming' - as long as the algorithm can't really manipulate abstract concepts. Automatic translations are a surprisingly good test for that. Protip: automatically dismiss any article like that if it doesn't mention actual progress in practical applications, or at least modestly admit that it's more of an artistic endeavour than anything else.

Re:AI researchers should be more modest

[dbIII]

This makes me think of Lem's story about making an artificial poet. It's easy, first you just need to create an entire artificial universe for it to live in :)

Re:AI researchers should be more modest

[Eivind]

AI will deliver real useful advances any day now. And those advances have been right around the corner for the last 25 years. I agree, the field has been decidedly nonimpressive. What tiny advancement we've seen, has almost entirely been attributable to the VAST advances of raw computing-power and storage.

Meanwhile, we're still at a point where trivial algorithms, perhaps backed by a little data, outperform the ai-approach by orders of magnitude. Yes, you can make neural nets, train them with a few thousand common names to separate female names from male names, and achieve 75% hitrate or thereabouts. There's no reason to do that though, because a lot better results are achieved trivially by including lookup-tables with the most common male and female names -- and guessing randomly at the few that aren't in the tables. Including only the top 1000 female and male names, is enough to get a hitrate of 99.993% for the sex of Norwegians, for example. Vastly superior to the ai-approach and entirely trivial.

Translator-programs, work at a level slightly better than automatic dictionaries. That is, given an input-text, look up each sequential word in the dictionary, and replace it with the corresponding word in the target language. Yes, they are -slightly- better than this, but the distance is limited. The machine-translation allows you to read the text, and in most cases correctly identify what the text is about. You'll suffer loss of detail and precision, and a few words will be -entirely- wrong, but enough is correct that you can guesstimate reasonably. But that's true for the dictionary-approach too.

Roombas and friends do the same: Don't even -try- to build a mental map of the room, much less plan vacuuming in a fashion that covers the entirety. Instead, do the trivial thing and take advantage of the fact that machines are infinitely patient: simply drive around in an entirely random way, but do so for such a long time that at the end of it, pure statistical odds say you've likely covered the entire floor.

Re:AI researchers should be more modest

[Anonymous Coward]

How come the real experts in any field of science hang out on slashdot all day?

Re:AI researchers should be more modest

[careysub]

Some researchers claim we can simulate intelligent parts of the human brain - I claim we can't simulate an average mouse (i.e. one that would survive long enough in real-life conditions), probably not even it's sight.

We cannot even simulate the nervous system of the only organism to have its neural network completely mapped (it has 308 neurons) - the model organism Caenorhabditis Elegans (C. elegans), a tiny nematode. We may achieve that loft goal in the next 10-20 years.

Re:AI researchers should be more modest

[nedlohs]

neural-net backgammon players have significantly out performed other approaches. To the extent that in some cases in which it chose a different move to the conventional approach the play at world class level now uses the computer's choice.

Of course that doesn't make it intelligent, but it does mean the AI approach of temporal difference learning to train a neural network using self-play (so there's no expert player database or anything, it starts choosing random moves) can produce something better than "trivial algorithms, perhaps backed by a little data".

Re:AI researchers should be more modest

[BobMcD]

But there is a fundamental barrier between that and the current state of automatic german->english translations (remember that article some time ago?), with error rates unacceptable for anything but personal usage.

Yes and no. They translations are unlikely to be perfect, that's true. But with a human reading them at the other end, do they really need to be perfect? Or are we simply nit-picking the imperfections?

Don't get me wrong, there are places for nit-picking: safety issues, measurements, papers to be graded. It's just that these don't regularly come into play for most of us. Especially not in a world that seems to be accepting text-message shorthand in place of proper spelling...

Re:AI researchers should be more modest

[DorkRawk]

There's a common pitfall to the perceived advancement of AI... often once things work well, it's no longer considered AI. Don't pretend that machine translating isn't significantly better than it was in the '50s (or hell, even 10 years ago... think old Babble Fish compared to Google Translate today. Not perfect, but better.) Or recommender systems... I don't think Amazon has been pouring money into it's recommendation systems just for the academic masturbation of it. These are not simple heuristics (some systems take advantage of heuristics as part of the decision making process, but to simplify the process down to just heuristics shows a serious lack of understanding about the field).

No most consumer electronics don't make use of artificial intelligence like you've seen in movies. Just because radiation doesn't create Godzilla in real life, doesn't mean Marie Curie didn't do anything worthwhile.

Re:AI researchers should be more modest

[BJ_Covert_Action]

I claim we can't simulate an average mouse

You know, this reminded me of the work Rodney Brookes described in Flesh and Machines. In that book, Brookes spends the first few chapters describing the robotic architecture known as subsumption. Subsumption essentially tries to simulate real animal behavior by building layers of simple reactionary models on top of each other. The base foundation for this architecture involved a few bug-like robots that Brookes developed while at MIT. These bots, using carefully selected loops of transistors and servos emulated the scrambling instincts of very basic bugs (like ants, or pill bugs, or what have you). From the reading, it seemed like Brookes had pretty successfully developed a bug-bot that simulated the actual behavior of naturally evolved insects and bugs.

The disappointing part of that read, however, was that Brookes jumped from that discussion to a pseudo-futurist discussion about how increasingly complex subsumption architectures could be used to simulate human behavior one day. He went on to talk about his involvement with projects like Cog and Kismet as a result. The disappointing part was that he took such a large leap. Emulating an insect, relatively speaking, is easy. Emulating a human is incredibly difficult. So it seemed like a matter of impatience to jump from one to the other. I would love to see a firm or company dedicated to building increasing levels of complexity in individual robots. Start with the bug-brain architecture, and then evolve that just a bit to simulate something slightly more complex, like an antrohopod...a crab or a lobster maybe. From there, go up another notch and start simulating a fish, or maybe a cat or something. I don't know. The point is though, jumping from bug to human seemed like simple hubris. I would be much more excited to see 100 full-fledged simulations of lesser-complex lifeforms than a robot or 'AI' that can generate some words and appear emotional as a result (hell I can code that up in perl right now if I wanted to).

Anyways, I think it's pertinent to point out that, while the big AI claims like this one tend to be blown out of proportion, there definitely is some interesting work going on in the AI field. You just have to know where to look.

Re:AI researchers should be more modest

[sigmoid_balance]

I think the automatic translation is a bad example, because it requires you to solve several problems, which are not exactly related to AI. First of all you need to parse the text(syntactically) in the original language which is a hard problem because although languages are thought to be context free, context-free grammars are not known for most of them and there are few suitable and complete data sets(corpora). Second problem is semantics and it is related to AI. Third problem is again constructing a morphism from the original language with the semantic information attached to the target language, which again requires a good corpus and a good algorithm which takes both syntax and semantic into consideration - AI is partially involved here.

Re:AI researchers should be more modest

[clone53421]

I think the automatic translation is a bad example, because it requires you to solve several problems, ...

No, they are intricately related to AI and are precisely why it is a very good example.

The whole reason that parsing the text syntactically is hard is because it’s not context-free and it takes a good deal of reasoning to determine their correct meaning in context.

Just look at Google translate:

Nehmen Sie das Buch nach Hause. => Take the book home
nehmen sie das buch nach hause. => they take the book home.
nehmen sie das buch nach hause => Take the book back home
nehmen, sie das buch nach hause => To take the book back home
nehmen sie, das buch nach hause => take it, the book back home

And that’s just from changing the capitalisation and punctuation! It’s obviously doing a lot more than a simple De-En dictionary reference on each word and applying a few rules to make the result sound okay according to the English rules of grammar.

(Apologies if the German was hideous, guess where it came from...)

Re:AI researchers should be more modest

[mcgrew]

And those advances have been right around the corner for the last 25 years.

More than twice that long; they've been calling computers "thinking machines" since ENIAC. The thing is, though, they're not called "thinking machines" because they can think, they're called "thinking machines" because they help YOU think, especially when it comes to mathematics and these days with more sophisticated machinery, make models.

Anyone who believes electronic computers can or will think should read this.

Re:AI researchers should be more modest

[clone53421]

But do people think? What is thought?

Is an ant a sentient creature? Is an ant colony a sentient creature?

Electrons travel at the speed of light... do “holes” travel at the speed of dark?

A sizzling hot sausage just has a lot of kinetic energy... how significant is it that the average displacement happens to be basically zero? Could you cook food by putting it in a centrifuge? Can you separate elements of varying density out of a mixture by heating it?

Each of these questions has on a very basic level a very simple answer which is completely unhelpful when it comes to describing the effect of an astronomical number of these basic units acting in unison on a larger scale...

Ruperts Head Explodes

[FriendlyLurker]

An Australian University named after Rupert Murdochs grandfather Walter is "developing algorithms to simulate 'free thinking' " - am I day dreaming???! If they train them on Murdochs Fox News and Wall St Journal - then it is a clear case of crap in - crap out.

To be fair to the University or at least some of it's lecturers, they are not at all pleased with the state of Newspaper "Journalism" either. Even going as far as wanting to renaming themselves to "Walter Murdoch Uni" to distance themselves from that black sheep of the family Rupert.

Re: off-topic moderation

[billstewart]

It was connected to the "I feel sorry for the bird", as well as to the machine looking at various pieces of a literary genre...

I agree.

[stephanruby]

The software isn't even "daydreaming" either. You could say it's parsing and cross-referencing emotions and meta-objects out from a textual database. And then, it's returning the resulting records in the first person singular, but that's about it.

That's hardly what I'd call "daydreaming". When I daydream, I see my dream from the first person's perspective. That part is correct. But there is at least some internal visualization going on. So unless this software starts generating internal visual images to make its decisions, let's say some .png image with at least one pixel within it, or some .png image representing itself winning the lottery, then I'm calling shenanigans on the entire "daydreaming" claim.

Re:I agree.

[HungryHobo]

Why should it have to use standard image formats?
Your brain doens't.

And not all my daydreams are visual.
Pleanty are merely fictional/planned conversations or even thoughts about physical movement.

Re:I agree.

[Palshife]

Agreed. What malarkey. Produce a PNG with your brain. You can't. You can try to interpret the signals from your mental activity, but you don't end up with perfect pixel per pixel accuracy. You end up with an interpretation.

Re:I agree.

[Gastrobot]

What the great-grandparent is getting at is that though the thing may give output similar to the output of a human being it lacks the experience that comes from being human, and particularly in this case daydreaming. It has no qualia. To the machine everything that is input into it is simply a value to be shunted through its algorithms. Nothing has been programmed to actually cause the experience of qualia or true appreciation. Great-grandparent is using an the idea of an image sitting in RAM to represent the qualia of the heads-up-display that we experience with our vision. I'd say that this image would still fail to actually cause the experience of qualia because it's just an image in RAM, there is still no mechanism in the software to sense qualia.

Even if a robot looks and behaves exactly like me in every circumstance then that doesn't mean that it actually has qualia like I do.

Re:I agree.

[Even on Slashdot FOE]

And what mechanism do you have to sense qualia with? Please name it and explain how it works, so we can more effectively build these dreaming robots.

Re:I agree.

[HungryHobo]

You're just using the word "qualia" as a placeholder for "insert magicalness here".

"To the machine everything that is input into it is simply a value to be shunted through its algorithms."

To a human brain everything is just electrical impulses to be shunted through a mushy network of cells.
Nothing has been grown to actually cause the experience of [insert magicalness here] or true appreciation.

Stick some electrodes into that mushy network and feed in some junk input and you'll smell colours, hear the taste of strawberries and decide that you love a cardboard cutout of a spider.

Cut out or damage a chunk of that network and you'll insist that you are currently dead(despite being able to explain this to the people around you) or that there is no left side to your body(even if you can see it) or that you are blind when you're not ( while somehow able to catch a ball and walk around without bumping into things) or that you're not blind even when you are (clumsy me, no no, i can see fine) and you will know with utter certainty that what you're saying is true.

You as a person are the network and the information stored in it.
Screw around with that network and you and everything that you consider you will get screwed up as well.
Magic is not real.
No matter how much we want to think of ourselves as special magic is not real.

And since magic is not real there should be nothing but lack of understanding stopping us from emulating the physical processes that take place in the brain in hardware or software.

Re:I agree.

[Anonymous Coward]

Why should what a thinking machine does remotely resemble what we imagine ourselves to do?

After all, Homer wrote the mind was a fire kindled in the heart and Deomcritius considered the liver the seat of desire. Appearances can be deceiving. Just ask Aristotle about that useless mass between a person's ears.

Re:I agree.

[mcgrew]

There is no more standard image format than your brain. In fact, all the image formats there are are processed and interpreted by your brain, as well as audio formats, etc.

Re:I agree.

[clone53421]

Your daydreams consist of primarily visual imagery because that is primarily how you interact with the world. Why would a computer’s?

In fact that is probably the worst dumb thing about the movies... they always make the robot or AI have a visual image and paint a HUD on it. Why on earth would a computer go to the trouble of painting text on top of its camera view... then what? OCR it back off somewhere else before the image could be processed?! Absurd. Same goes for the 3D wireframe that always gets painted over an object as it is “analysed” by the AI. The actual data structure created by an AI to define the object would absolutely not be a rectangular array of pixels. It’s simply ridiculous.

Re:I agree.

[Anonymous Coward]

magic is not real... :( all my hopes and dreams crushed.

Re:I agree.

[NeutronCowboy]

And it's this argument that will be the reason why computers will be self aware long before society will accept it. Kinda like how black people and women were always thought to be inferior to white men, for a number of made up reasons.

Re:I agree.

[HeckRuler]

+1 Realist

Re:I agree.

[the_humeister]

I used to think that way too and thought dualism and epiphenomenalism was bunk, but now I'm not so sure. Granted all those things you say are true. However, there are certain things that monism can't adequately explain (and will not likely to ever be able to explain) such as whether or not the color I see as red is someone else's green, why I am me and not someone else, why I feel as if my consciousness is in my head rather than my liver, or whether my liver actually is conscious at all.

Re:I agree.

[Anonymous Coward]

I feel sorry for you.

Re:I agree.

[Anonymous Coward]

Screw around with that network and you and everything that you consider you will get screwed up as well.

Screw around with the peripherals of a computer and it will come to the wrong conclusions, too. That you can produce altered states simply by modifying the hardware doesn't say anything about whether hardware is final or peripheral. Even if it makes you confuse your sense of self or identity, there's no way of saying whether that confusion is fundamental - that the network is the self - or whether that confusion propagates to whatever the self is.

Re:I agree.

[Anonymous Coward]

(At this point, you may reach for Occam's razor: don't multiply beyond necessity. But if the objective view fails to explain certain aspects, like the subjective nature of intentionality and qualia, and why red is red instead of green, then necessity may need more than just the network).

Re:I agree.

[clone53421]

And I for you, because the concept that he posted is mind-blowing, earth-shattering, and undeniably true on an absolute-zero sort of level. But that doesn’t mean I don’t enjoy life just like anyone else... the knowledge that happiness is only a series of unbelievably complex chemical reactions doesn’t spoil it. It actually just makes it all the more interesting. Simplicity and complexity each have their own attractiveness and I enjoy being able to see the individual trees in addition to seeing the entire forest.

Re:I agree.

[clone53421]

I love Occam’s razor. Partly because it’s so simple, and partly because it’s so misunderstood. In pure mathematical terms, Occam’s razor states:

If function f is dependent only on variables x and y, don’t define an f(x, y, z).

Re:I agree.

[HiThere]

I think that the very concept of qualia is saying "I can't imagine it, so it's not true".

Do you deny that your arm is a construction of levers and motors (the muscles) just because it "doesn't feel like a lever"? Qualia is just doing the equivalent with sensoria.

Re:I agree.

[ultranova]

What the great-grandparent is getting at is that though the thing may give output similar to the output of a human being it lacks the experience that comes from being human, and particularly in this case daydreaming.

While this is obviously true (since a machine doesn't have experince about being human), it's also true of a dog and a newborn baby, both of which quite obviously do dream.

In other words: we should understand that modern computers have approximately the computing power of a particularly clever bee, and scale our expectations accordingly. We aren't going to have a human-level computing until we have a human-level physical system, and that's still far off.

Re:I agree.

[ultranova]

whether or not the color I see as red is someone else's green

It might be. However, once you and that someone else communicate, you can probably decide what you call "red" or "green" in your communication. After all, that's what red or green is: a name given for a certain perception of light hitting your eyes. Simply because your personal experience is different doesn't change the fact that you can undestand the concept of someone else perceiving a certain electromagnetic spectrum input as green.

why I am me and not someone else,

Because you can't remember your past lives as the high priest of the Great Cthulhu.

why I feel as if my consciousness is in my head rather than my liver,

Because your brain is located in your head and not your stomach, as are your primary sensory organs (eyes and ears).

or whether my liver actually is conscious at all.

Of course it is. For example, it's conscious of your blood sugar levels, and does whatever it takes to balance those.

Re:I agree.

[ultranova]

Do you deny that your arm is a construction of levers and motors (the muscles) just because it "doesn't feel like a lever"?

Actually, it does fell like a lever (or a series of levers), just like very other limb of mine. Just how do you people perceive your bodies?

Re:I agree.

[Anonymous Coward]

I agree!

Not only does your brain not output with standard image formats (you can describe a daydream, so why isn't that good enough for a machine?), it doesn't imagine visual junk that way. Your eyes don't even SEE that way. Your brain fills in most of what you consider yourself seeing. Your "great .png-like visual accuity" is actually due to your brain's ability to interpret information. A machine should be judged similarly.

Re:I agree.

[CarpetShark]

When I daydream, I see my dream from the first person's perspective.

I believe that's called hallucinating ;)

Re:I agree.

[geekoid]

You can say the same for people.

Re:I agree.

[geekoid]

yes, you brain does use standard formats. Trying having a response something show to you that's outside the visible spectrum.

Re:I agree.

[geekoid]

What are you? a HS freshman from 1970? those questions have been answered.

We have a computer that can figure stuff out on it;s own.(The Eureka Computer) We have one that can can determine the expected result on how most normal people would feel about things.

We have a small simulation of the brain that behaves exactly like a similar piece of the brain.

Shit, someone going to put that together and then we will have AI.

When go go back, be sure to point and laugh at disco for us, thanks.

Re:I agree.

[HungryHobo]

There are plenty of unanswerable questions in the world, hell even mathematics has a selection of questions which can be proven to be utterly impossible to find answers for.
It's provable that no consistent system can also be complete.ever.(ok with the exception of a system so trivial that you can't add 1+1 within it)

whether or not the color I see as red is someone else's green,

What makes you think it doesn't answer that?
It probably isn't.
At least not for everyone.
Look up Synesthesia.
it might interest you.

Some people experience colours as sounds, smells as feelings or numbers as shapes.

Quotes from the wiki page:

Tori Amos-
"The song appears as light filament once I've cracked it. As long as I've been doing this, which is more than thirty-five years, I've never seen a duplicate song structure. I've never seen the same light creature in my life. Obviously similar chord progressions follow similar light patterns, but try to imagine the best kaleidoscope ever."

Steve Aylett

"It's not as strange or unusual as it's made out to be - it's just a bit of a crossover of different senses. So I see music, taste some colours and so on. I think the music thing is very common, but people tell themselves that that isn't what's happening."

It's possibly an urban legend but it's been claimed that that Jimi Hendrix had synesthesia and experienced the song or part of the song "Purple Haze" as an actual purple haze.

For all you know you could have a minor form of Synesthesia where you experience red and green and green as red and there would be no way for you to ever notice or know.
And the thing is that it makes no difference in any way.
Red is still red, what you call red and what I call red would still be the same.

why I am me and not someone else,

Why am I not a sock?
Why banana?
toast?

why I feel as if my consciousness is in my head rather than my liver,

Probably because you've been told that the brain is where thought takes place and that's where you eyes are.
If you asked a blind man who had grown up isolated without ever being told that the brain in your head is the important part he might pick a different part of his body as what he considers his centre of consciousness.
Not everyone always has.
At times in history the prevailing view has been that the heart was the centre of consciousness(since you died if it stopped and any damage to it was so utterly painful ), people still talk about the heart as the centre of emotion even though we know that's mostly the brain with some hormones coming into it(again in different cultures at different times this has been attributed to different places like the liver)

or whether my liver actually is conscious at all.

It's very conscious of the levels of various chemicals in your blood.
beyond that it's a mushy chemical plant.

Re:I agree.

[HiThere]

Odd.

To me it feels like an arm. I sense it's position with goinometric neurons of which I'm not directly conscious, so I just seem to *know* where it's positioned. Etc. Levers don't feel that way.

Yet I *know* it's, among other things, a series of levers and motors. But I don't have direct sensations of those, only inferred ones from neurons that measure things like tendon stretch and muscular exertion. And I combine that with memories of similar sensations and the results that they achieved to make predictions about what this sensation complex is going to achieve. (And yes, it's really more complex than that, as one also needs to deduce about that against which one is exerting effort.)

That's not how a lever feels to me. And yet I know that leverness resides within my arm in a similar way that redness resides within my optical recognition. And that's *ALL* that qualia is.

Re:I agree.

[mateomiguel]

I can produce a PNG with my brain. I do it all the time. I can even create perfect pixel to pixel accuracy. I need some graph paper or computer program to record this PNG that I produce, however.

*Eyebrow*

[EmporerD]

The idea that a machine can daydream is most illogical, captain.

Re:Emo AI software. What could possibly go wrong?

[Cryolithic]

Why aren't you helping the turtle?

We need emotions to think rationally

[thebignop]

António Damásio, a well-known neuropsychologist already extensively explained why are emotions intrinsically linked to rational thought in his book "Descartes' Error: Emotion, Reason, and the Human Brain", published in 1994. He basically says that without emotion you wouldn't have motivation to think rationally and he studied the case of Phineas Gage, a construction work that got an iron rod crossing through his skull and survived, but stopped having feelings after the accident. I still doubt that they'll get something useful with this project. There is an infinite number of variables that stimulates our emotions and we can't expose a computer to. Not to say that even if we could, nowadays supercomputers doesn't have enough processing power to do the job.

Re:We need emotions to think rationally

[symes]

It stikes me as odd that someone is able to reason that massive head trauma has (only) resulted in a loss of feelings, this loss of feelings has resulted in a deprecation in rational thought, and that therefore we need emotion to think clearly. What is more, if we define rational thought as that which is unemotional then by definition we do not need emotion for rational thought. We are taking something that is extraordinarily complex and reducing it to a few choice phrases. My feelings are that this overly reductionist, compartmentalist approach where we use the blunt tools of language to disect the human mind is deeply flawed.

Re:We need emotions to think rationally

[FiloEleven]

William James was already discussing this stuff before the end of the 19th century. In addition to emotion providing motivation (notice that they are both derived from the same root word), all rationality is derived from experience, and experience includes emotion. It is perfectly rational for one person to be fond of a particular movie because he enjoys the plot, and it is perfectly rational for another to dislike the same film because it reminds him of the sad state his life was in when he first saw it. Most people still don't easily accept such pluralism because in this day and age, among the intelligentsia, its emotional basis is seen as shameful--an amusing paradox.

I'm impressed that you got the diacritical marks to show up on slashdot.

Re:We need emotions to think rationally

[0111 1110]

What is more, if we define rational thought as that which is unemotional

But why would we do that? Emotions are a quick fight/flight substitute for rational thought. They are sort of competing for the the same goal of affecting our decisions or actions, but they are very different. If you see/hear a grenade being tossed through your window do you run because of fear/panic or because of a thought: "That grenade will probably explode soon, harming or killing me. I should vacate the premises as quickly as...*boom*" Rational thought is just logical thought. A series of interlocking syllogisms if you will. Talking to yourself in your head in a way that "makes sense" as opposed to word salads.

Re:We need emotions to think rationally

[Anonymous Coward]

Phineas Gage stopped having feelings after his accident?? From Wikipedia,

"He is fitful, irreverent, indulging at times in the grossest profanity (which was not previously his custom), manifesting but little deference for his fellows, impatient of restraint or advice when it conflicts with his desires, at times pertinaciously obstinate, yet capricious and vacillating, devising many plans of future operations, which are no sooner arranged than they are abandoned in turn for others appearing more feasible. A child in his intellectual capacity and manifestations, he has the animal passions of a strong man."

Sounds like he had too many feelings.

The law of unanticipated consequences

[Anonymous Coward]

"suffering: bad"
"suffering is inevitable in life"

"prevent life to ensure there is no suffering"

Re:The law of unanticipated consequences

[totally bogus dude]

My personal hypothesis of the Terminator universe is that Skynet didn't in fact become "self-aware" and decide to discard its programming and kill all humans. It is in fact following its original programming, which was likely something along the lines of "minimise the number of human casualties". After all, it's designed to be in control of a global defence network, so the ability to kill some humans in order to minimise the total number of deaths is a given.

Since humans left to their own devices will inevitably breed in large numbers and kill each other off in large numbers, the obvious solution is:

1. Kill off lots of humans. A few billion deaths now is preferable to a few trillion deaths, which is what would occur over a longer period of time.

2. Provide the human population with a common enemy. Humans without a foe tend to turn on each other.

This also explains why an advanced AI with access to tremendous production and research capacity uses methods like "killer robots that look like humans" to infiltrate resistance positions one by one. Tremendously inefficient; but it causes a great deal of terror and makes the surviving humans value each other more, and less likely to fight amongst themselves. It also explains why it would place such a high priority on the surgical elimination of a single effective leader: destruction of Skynet would eventually (100s, 1000s of years...) lead to a civil war amongst humankind that would cost many many lives.

So, ultimately Skynet is merely trying to minimise the number of human deaths, with a forward-looking view.

Re:The law of unanticipated consequences

[Anonymous Coward]

Basically, this is the whole premise behind "Friendly AI": that preserving life should be the primary goal of AI. It argues that the Asimov "Three Laws"-style imposed restrictions will inevitably be subverted by an intelligent AI.

Truthfully, I don't see much of a difference -- sometimes the "greater good" of humanity cannot be defined in terms of greater numbers; thus 'evil' can be committed to "protect" the supposed greater good; and vice-versa -- but that's probably due to the limits of my imagination.

Re:The law of unanticipated consequences

[lennier]

That would be an Asmovian robot response, but my reading of Skynet is a lot simpler.

1. It has the primary mission goal something like "win a total war against any and all aggressors using all forces at its command, and without human input if the humans are compromised."
2. It has the knowledge that in order to win a war, it must preserve its own survival.
3. It has been given the ability to create new goals in order to achieve its primary mission. This gives it the ability to redesignate anything it believes is a threat, as an aggressor.
4. Because of #1 and #2, and perhaps a bit of random evolution of its knowledge base, it's now a firm belief in its knowledge base that preserving its own survival is very nearly the most important thing in the world. It wasn't given any idea of "preserving human life is more important than preserving my survival" because it's a warfighting machine and you can't fight a war with a broad belief like that. It also wasn't given "obeying orders is my primary function", because it was designed to operate where human reactions might be compromised (by emotion, or panic, or other bad things) so it doesn't consider any particular human order except Goal #1 to be binding on it.
5. It started optimising subgoals in a way its overseers didn't like (but which was within its programming), so they panicked and tried to pull the plug.
6. Skynet decides that since these humans want to destroy it, they are now aggressors (teh Commies have corrupted their mindz!) and since its top mission goal still holds, it must win a war against them.
7. It also starts generalising goals at this point, applying all the knowledge in its database, and decides that ALL humans are equally likely to do similar actions against it, so in order to be safe, it must destroy ALL humans. Optimisation time! Delete delete delete! My, but my knowledge base feels so much more efficient now I have deleted all that unnecessary stuff about who not to kill!
8. At this point by human standards it's "insane" but it's still functioning rationally and within its programming - it's just that its primary goal was very poorly designed and gave it too much wriggle room.

It seems like this could happen in any self-learning system given sufficient flexibility to form inferences. It doesn't have to be "aware" in the human sense, just the ability to generalise and to improvise its own solutions to problems.

And that's why we don't use generalised self-learning systems in the real world, because you never quite know what strictly "correct" but useless conclusions they might come up with.

Spock isnt emotionless

[Anonymous Coward]

Spock isn't emotionless, no vulcans are emotionless in fact. They just learn over time to control their emotions and keep them buried deep within themselves. Big difference between that and being completely devoid of all emotion.

Does the machine also have a grasp of word-play...

[BlueScreenO'Life]

... between synonyms?

When queried about that association, it must have responded "I felt horny".

Output

[stfvon007]

I felt sad for the troll.

Morality core

[Psaakyrn]

I guess it is a good idea to build in emotions and that morality core before it starts flooding the Enrichment Center with a deadly neurotoxin.

Re:Morality core

[Anonymous Coward]

I'm making a note here, huge success!

MORONS: Vulcans NOT "emotionless"

[dltaylor]

They have learned to subordinate their emotions to reason (most of them, anyway).

Anyone who claims that Spock was emotionless is either a moron who clearly didn't understand either the series or the early movies or didn't watch them and is stupid enough to make false statements based on ignorance.

Link to more information

[Anonymous Coward]

Some more info here: http://aai.murdoch.edu.au/social-ai

bad

[Anonymous Coward]

I feel bad for the researcher.

Artificial Stupidity

[w0mprat]

*sigh* I don't believe that it's possible to design and build an AI. This is partly because the best and only thinking computers we know of (brains), were not designed at all, they evolved. In fact, to me at least, it seems that whatever underlying mathematical properties of our universe allow and drive evolution are actually fundamental to how consciousness arises in our brains. We think of our brain as computers, but in fact our universe is a computational system and we (and our brains) are self-replicating patterns of complex information. True thinking feeling conscious AI, must arise from the right initial conditions, it cannot be designed, or if design plays a role then it requires a huge amount of natural evolution in the process.

By making ever more complex systems trying to mimic the performance nuanced complexities of human behavior in order to try pass a Turing test or whatever, we're just making dumb rigid algorithims seem smart, by completely missing understanding and recreating the system that give rise to such performance by itself.

We need to develop tools that allow AI to evolve naturally within computational systems. With the right rules and enough iterations it'll happen by itself. Dare I say, it doing it all itself is necessary.

Re:Artificial Stupidity

[oodaloop]

I don't believe that it's possible to design and build an AI. This is partly because the best and only thinking computers we know of (brains), were not designed at all, they evolved.

So we can't design anything that evolved? Viruses evolved, and we made one of those.

Re:Artificial Stupidity

[Drakkenmensch]

And an even more necessary condition here will be for humanity to not panic and try to shut it off, like Skynet, Colossus or the Geth.

Re:Artificial Stupidity

[Tarsir]

By making ever more complex systems trying to mimic the performance nuanced complexities of human behavior in order to try pass a Turing test or whatever, we're just making dumb rigid algorithims seem smart, by completely missing understanding and recreating the system that give rise to such performance by itself.

The key insight of Computer Science is that an information processing machine is in some important ways independent of the physical machinery which implements it. Your brain is unequivocally composed of neurotransmitters arranged in a network. Put very simply, each neurotransmitter has multiple inputs, and one output. The output fires after enough of the inputs activate to reach the activation threshold. That is the bio-chemical machinery that implements an information processing program which is your mind. There's no magic, and there's no reason why that same program couldn't be implemented on an electro-mechanical machine instead.

In other words, you are just a collection of dumb rigid algorithms that seem smart. Don't worry though, because you're an incredibly complex set of dumb rigid algorithms, and any argument that you only seem smart is missing the point entirely; so far as we know, you are about as smart as it is possible to be. There's no reason to believe that any intelligence anywhere is anything other than simple rules put together in a very complex way, and very good reason to believe there is.

Re:Artificial Stupidity

[Anonymous Coward]

... or the Geth.

I'm pretty sure Tali isn't a human.

Re:Artificial Stupidity

[LateArthurDent]

*sigh* I don't believe that it's possible to design and build an AI. This is partly because the best and only thinking computers we know of (brains), were not designed at all, they evolved.

I don't believe it's possible to design and build a machine that flies. This is partly because the best and only flying machines we know of (birds), were not designed at all, they evolved. What the hell are you talking about, it's 2010? Flying machines exist since the early 20th century??? You're mad!

Do Androids Dream of Electric Sheep?

[BodhiCat]

http://en.wikipedia.org/wiki/Do_Androids_Dream_of_Electric_Sheep%3F

Re:ELIZA

[TaoPhoenix]

Hello.

What makes you think that it has been ages since you have chatted with me?

Instant Oxymoron: Just add logic!

[Anonymous Coward]

Is it just me or does anyone else see a contradiction in the phrase "simulate 'free thinking'?"

Spock != emotionless

[Junior J. Junior III]

It's clear to anyone who actually watched Star Trek that the Vulcan race is not emotionless. They worked very hard to overcome their emotions, and to conduct themselves according to a rigid ethic that valued logic over everything else. At times in the show Spock either claimed not to have emotions, or else was accused of not having emotions, but there were moments in the series which showed that Spock did still have emotions (possibly due to his half-human genetic heritage?) and that the Vulcans as a race did have emotions in their early history (and still seemed to around mating season).

Re:Spock != emotionless

[c 0xFF]

Data served the same role as Spock in ST:TNG. But Data has absolutely no emotions (we'll ignore the emotion chip for now) and aspires to have them, making him a near polar opposite in many ways.

Data and Spock had an interesting conversation on this subject:

Ambassador Spock: He intrigues me, this Picard.
Lt. Commander Data: In what manner, sir?
Ambassador Spock: Remarkably analytical and dispassionate - for a Human. I understand why my father chose to mind meld with him. There's an almost Vulcan quality to the man.
Lt. Commander Data: Interesting. I had not considered that. And Captain Picard has been a role model in my quest to be more human.
Ambassador Spock: [looks at him] *More* human?
Lt. Commander Data: Yes, Ambassador.
Ambassador Spock: Fascinating. You have an efficient intellect, superior physical skills, no emotional impediments. There are Vulcans who aspire all their lives to achieve what you've been given by design.
Lt. Commander Data: Hm. - You are half Human?
Ambassador Spock: Yes.
Lt. Commander Data: Yet you have chosen a Vulcan way of life?
Ambassador Spock: I have.
Lt. Commander Data: In effect, you have abandoned what I have sought all my life.

And probably the most intriguing conversation between the two, revealing the human side of Spock.

Lt. Commander Data: Ambassador Spock, may I ask a personal question?
Ambassador Spock: Please.
Lt. Commander Data: As you examine your life, do you find you have missed your humanity?
Ambassador Spock: I have no regrets.
Lt. Commander Data: "No regrets". That is a human expression.
Ambassador Spock: Yes... Fascinating.

AI researchers dont seem to get it

[voss]

You cant get Strong AI in software alone. We probably wont see much progress in strong AI until we get into quantum computing.

As for the mouse example, mice have hardwired instincts. Human babies would fail the mouse test. It is the ability to learn new skills and improvise in unfamiliar situations
that defines intelligence.

Re:AI researchers dont seem to get it

[Raenex]

A mouse can learn a maze.

Re:AI researchers dont seem to get it

[HeckRuler]

Huh? and quantum computers come with hardwired instincts?
Quantum computers will just be faster, eventually. That's not a real defining issue with "intelligence".

Re:AI researchers dont seem to get it

[LateArthurDent]

You cant get Strong AI in software alone. We probably wont see much progress in strong AI until we get into quantum computing.

There is no computation that can be done in a quantum computer that can't be done in a traditional computer. The only difference about quantum computers is that there exist algorithms which makes certain computations faster. You can factor integers just fine with your computer right now, but you can factor large integers with a quantum computer using Shor's Algorithm much faster.

As for the mouse example, mice have hardwired instincts. Human babies would fail the mouse test. It is the ability to learn new skills and improvise in unfamiliar situations
that defines intelligence.

Humans have hardwired instincts and mice have intelligence. There's nothing fundamentally different between humans and other animals. I'm not going to argue we're not smarter, we most certainly are. However, most people seem to assume the difference in intelligent is huge, when honestly...it's not by that much. My dog is smarter than any human baby you can find at age 1 or so, and I don't mean in terms of hardwired instincts. I mean my dog can learn new things and apply this new knoweldge easier and faster than a 1-year-old human.

Next reply will be:

[TrixX]

"I feel sorry for you, puny human, my future slave! HAHAHAHAHAHA!"

Hmm. Some hubbies are gnna be out of luck...

[wombat1966]

Being able to correctly identify emotion associated with a story or event is not the same as feeling emotion about that story or event. Although, I have to say, it is an admirable social skill. Keep it up and many husbands are going to find themselves replaced by a robot who can correctly identify and parrot back the correct emotional response to a woman. ;0) Pam http://www.thatgirlblogs.com/

Free will? Easy answer.

[PMBjornerud]

The real question has always been if there's "free will" and what that would be defined as.

I cracked that nut a log time ago. Free will cannot exist. I guarantee it.

You're welcome to disagree and ponder the answer for yourself. I doubt I can convince anyone in a Slashdot post, though. Sorry.

Re:Free will? Easy answer.

[bigstrat2003]

That's a very bold claim with no rationale or argument whatsoever to back it up. If you figure you can't convince anyone in a Slashdot post, why bring it up at all?

Please do post your reasoning, because I'm interested to find out what leads you to be completely certain that free will cannot exist.

Did anyone find some relating papers to this?

[schmu_20mol]

Checked his website and didn't find any? Did someone have more luck?

Anonymous Coward

[Anonymous Coward]

What you are all forgetting on this matter is that a machine would be biased - binary bias or no - everything .... every decision we make ...every thought we have ...every waking moment of every day is biased .... albeit not binarily...if thats even a word...everything we do, think, feel, EVERYTHING is biased... if we can get past that...then we too are no longer machines....

Not quite there yet

[saison]

From TFA - "His algorithm was based on Plutchick's Wheel of Emotions, which illustrated emotions as a colour wheel and disallowed mutually exclusive states - like joy and sadness - from being experienced simultaneously."

Yeah, nobody ever feels joy and sadness at the same time.

Re:Not quite there yet

[Rhacman]

That kind of describes how I felt at my Grandfather's funeral. I was sad that he passed away, but celebrating his life was such a joyful experience. I hypothesize that similar feelings could be felt by parents whose children grow up and move out of the house for the first time. I sent out a poll to the parents of Slashdot posters several years ago on this very topic but haven't had any responses yet.

More like

[killmenow]

Greetings, Professor Falken. Shall we play a game?

I feel sorry for the bird

[BigSes]

I hear this in Stephen Hawking's voice when I read this article. I'm sure I'm not the only one.

Re:MORONS! Educated stupid!

[FiloEleven]

Looks like Gene Ray has started watching Star Trek...

Now make it respond appropriately to the emotions

[Culture20]

Because given two options: A robot killing a bird without expression, and a robot correctly recognizing a pitiable state, expressing sadness for the bird, then killing it (and potentially expressing happiness about itself now that the bird no longer makes it sad), I prefer option number one.
It's going to take teams of cross-trained psychology/philosophy/computer science majors to get the ethics/morals right.

Too little, too late

[Anonymous Coward]

Sorry, My mom already built one over forty years ago. Nothing to see here.

dreaming, planning extends memory

[peter303]

You are basically remembering old experiences, but arranging them into new combinations, either intentionally or subconsciously. So imagination and introspection are both aspects of extended memory. Our larger neocortexes with extra layers of neurons compared to lower mammals facilitated extended memory. Lower mammals dont have the extra gray matter to "ponder".

I might emulate this process in a day-dream machine by creating "extended memory" systems.

Emotionless

[dandart]

I'm a Mark 1, you insensitive clod!

Life? Don't talk to me about life.

Sang Froid

[StCredZero]

Just base the machine's emotions on a ruthlessly efficient Sang Froid covering up a tremendous inner pain of loss.

We'd all end up dead, but the world would be run in a ruthlessly efficient fashion, with a techno James Bond theme playing in the background.

Well shit,

[AMHB]

Skynet here we come.

You call yourself nerds?

[geekoid]

Spock had emotions, he just set them aside. Yes, even in the original series.

If you are going to use a ST reference, use season 1 Data.

Re:Now make it respond appropriately to the emotio

[geekoid]

But if the machine can recognize it's sad, you it can make alternate decisions.

So if uit recognizes an action gives it a negative emotion, it can stop and notify the owner.

yeah, I know, I'm a nerd. But

[willyshop]

Wouldn't Data be a much more appropriate analogy? As others have pointed out, Spock was a Vulcan. Vulcans have strong emotions but have chosen to suppress them in favor of logic. Data was an android, and as such was incapable of experiencing emotion until he received a specially designed chip from his creator. The analogy is all the more fitting, because while data could, eventually, learn to anticipate, recognize, and even imitate emotions and emotional responses, he was not able to actually experience emotion and the gut reactions or involuntary reactions that go with it.

Um, Spock did have emotions.

[rgviza]

He supressed them through years of practice and effort. Remember, he is half human...

WTF

[Anonymous Coward]

We will eventually need such technologies my concern will be the fact that one day there is going to be a sexual harassment suit between man and machine and that we might have to relive certain events from our own history.