Again with the Robots.

Do people have an opinion on that thought experiment that’s usually called the “Turing Test” or “Imitation Game”? You know, the one where we have a person, a machine and an interrogator. And the interrogator interacts with both the person and the machine via a monitor (e.g. chatting). The interrogator has to guess which is the person and which is the machine. If the machine can convince the interrogator that it itself is the person then the machine should qualify for “intelligence,” or “human intelligence,” or maybe “personhood” or something.

perhaps intelligence- otherwise i think by definition you just said mentally handicapped (not sure if i'm aloud to say that out loud) people aren't "persons"...

If we say the machine is a person it doesn't necessarily, I think, mean that mentally disabled people wouldn't be persons—but I get your point—because it could say that if we defined persons as those that only exhibit behaviour that convinces us of their human-ness.

BUT, regardless, even if we did say that these machines were persons and that persons were only those that had these qualities, that still wouldn't mean that mentally handicapped people should be stripped of their human rights, not cared for etc.

But, yes, thank you for the response. And now I'll try push the question a little further and ask "Okay, we think it likely that this machine has intelligence. Now, do we think it a 'person', or something that should be afforded some sorts of 'rights'?"

No.

Scorpio wrote:No.

Second

It's a machine- and no i don't think we've developed machines to the point where they can think for themselves... although i think this guy is slowly becoming self aware:


My personal opinion (which i am sharing with you all free of charge) is that we shouldn't worry about right's for machines until we've sorted out how we uphold basic rights for humans (which would be found here). If we can't look after our own (and i mean that as an entire Human race) then we shouldn't bother with anyone else, that sounds ordinary- but i'd rather feed a starving kid in uganda than soothe my washing machine's depression!

I absolutely second THAT!

Fuck the panda's! People first (as Chloe knows only too well)!

I think Trev and Tom's first responses could have been better summed up with this:

"There's a can."

Glenjamin wrote:It's a machine-

I believe we are too.

Glenjamin wrote:and no i don't think we've developed machines to the point where they can think for themselves...

Not yet, no... Hence the thought experiment, and the inexorable future.

Glenjamin wrote:is that we shouldn't worry about right's for machines until we've sorted out how we uphold basic rights for humans

Well, I think we kinda have; we're just no good at implementing it, and we possibly never will be. I think the thought experiment is supposed to try address whether we're able to overcome our self-important species-ism (evidently not), and whether we feel comfortable with extending basic civil rights to an entity that is capable of reciprocating those rights and expressing emotions/intelligence indistinguishable from humans (i.e is effectively another person). If it helps to soothe your anthropocentric urges you can imagine the robot with a similar human appearance.

I don't see the reason for 'getting over our species-ism'. We. Are. The. Best.

We just plain are. We may not be the 'best' at getting coconuts out of trees, or of breathing under the ocean, but we can make arm longerers and swim suits and air tanks. So eat that fishies!

Until a computer is truly, truly capable of independant thought to the point where it can ponder and even end its own existance on a whim then I will continue to hold all things, great and small, in low regard. Especially since i think that we should tackle things like 'racism' and 'classism' before we tackle 'hatred of almost thinking machines'.

I did a course in university which was all about possibilities of other intelligent life existing anywhere else. this included an entire lecture on the possibility of silicon lifeforms or AI. At the time of hte lecture (2003) the average home PC was as 'smart' as a grasshopper, meaning that the amount of processes it was capable of were the same as a grass hopper. By 2020 the average home PC will be capable of the same processes as a human being (atm we about 5 million times better than a computer). By 2040 (this is going on the current trend in computer technology which i believe may have slowed being that they are now just packing in more processes rather than changing the way they do it) the average PC will be able to do the same amount of processes per second as the entire human race combined.

But even this would not mean that the machine was intelligent. The software would have to be self adaptive, meaning that in any situation, even situations not dreamt of by its creater it would have to be able to adapt.. A computer with an incredible piece of software, that has voice recognition and a massive data base of prepared responses is still a computer and a tool, even if it can fool a man in a room.

I'm going to have to agree with Tom- and disagree with this:

Andrew.C wrote:Well, I think we kinda have; we're just no good at implementing it, and we possibly never will be. I think the thought experiment is supposed to try address whether we're able to overcome our self-important species-ism

We haven't worked out HOW to uphold basic human rights at all- sure we've got them written down and the UN can give itself a huge warm and fuzzy over that, but while we're staring lovingly into our own navels and telling ourselves what fine policy makers we are the following occurs:
General Human rights abuses according to Amnesty International

Crisis in the DCR

Genocide in Sudan

Do i have to keep going? I really with this idea that "we're doing ok, not great but fine for now" mentality- millions of people being butchered for political gain is not fine, nor ok, in fact i doubt it's particularly sane. Before we even think tank an idea like this- why not spend some time on the real issues that humanity face... and if you're i'm poor, i can't help- this site opened my eyes a bit

Tom wrote:Until a computer is truly, truly capable of independant thought to the point where it can ponder and even end its own existance on a whim then I will continue to hold all things, great and small, in low regard.

Sure, but what about when (or "if", if you prefer) they can?

Tom wrote:But even this would not mean that the machine was intelligent. The software would have to be self adaptive, meaning that in any situation, even situations not dreamt of by its creater it would have to be able to adapt..

What if it can? Would that make a difference? Then let's suppose it can.

Tom wrote:A computer with an incredible piece of software, that has voice recognition and a massive data base of prepared responses is still a computer and a tool, even if it can fool a man in a room.

Okay, fair enough. So, just to get an understanding, are you saying that the specific game that I presented wasn't good enough to establish whether a robot is actually "intelligent," in the sense that we call ourselves that. And do we need a better thought experiment? Or is it that you're saying an 'artificial' machine (as opposed to a biological one) can never be thought of on the same level as a human?

Glenjamin wrote:We haven't worked out HOW to uphold basic human rights at all- sure we've got them written down [...]

Sure, sorry, that is what I meant. And I never said the situation was "fine" or even ok, and that wasn't my point.

Glenjamin wrote:Before we even think tank an idea like this- why not spend some time on the real issues that humanity face...

True... but I don't feel the two mutually incompatible or anything. On the contrary perhaps: Developing a sound basis for human rights that attempts (probably not a possibility) to remove all forms of prejudice is, I feel, the only way to achieving the ideal goal, and that development is a direct consequence of discussing these sorts of things. It also involves actually thinking about issues rather than just running head first into the situation unprepared expecting everything to work out fine. Implementing 'peace' all over the world is an almost insurmountable logistical problem, and all we can do is essentially talk about it, isn't it? So, dealing with these "real issues that humanity face" involves discussing the nature of rights and why we think certain beings have them, and, indeed, why we think we can divvy them out.

Second, the thought experiment most likely will be a reality at some point in time.

My bug-bear with this discussion (and yes i am more than happy to be involved ) is that it seems to suggest "we can't fix our current problems- lets fix other ones in the hope that these ones will sort themselves out!" which i'm sure you'll all agree is ludicrous... I wonder if racism isn't at the heart of this discussion? not that i'm saying Andy's original aim was to discuss this, but i think if there were no racism there'd be no worry about accepting super evolved machine (the sort that could really think for themselves... like this guy- ).

So why not focus on that- why is it that people can't shake racism, and yes i think EVERYONE has an issue with it, it's almost like we can't function without it. So yeah, i doubt we'd ever give "personhood" to an intelligent machine- seeing as we can't seem to give it to the people from other nations!

Glenjamin wrote: My bug-bear with this discussion (and yes i am more than happy to be involved ) is that it seems to suggest "we can't fix our current problems- lets fix other ones in the hope that these ones will sort themselves out!" which i'm sure you'll all agree is ludicrous...

Yes, I do agree, but I don’t think what I was saying did suggest that, personally.

Glenjamin wrote:So yeah, i doubt we'd ever give "personhood" to an intelligent machine- seeing as we can't seem to give it to the people from other nations!

Yeah, that could be true.

Glenjamin wrote:So why not focus on that- why is it that people can't shake racism,

Yeah, okay: I would say that it was human nature, unfortunately.

I have no problem with gifting machines with personhood. I also have no problem with the concept of a machine that can truly be sentient. However! We are no where near that mark yet, no where near. To even suggest that a computer today could be considered sentient is ludicrous.

I agree. Not today. Within our lifetimes? Maybe (likely). Within this century or the next? Inevitable (probably). Which is really a blink of the eye in terms of human development, not to mention the expansive time scale of life itself on this planet.

Giving rights to one group doesn't mean it prevents the giving of rights to another. There's no reason why we can't give rights to humans and machines simultaneously: apart from the massive logistical problems involved in, ya know, fixing all the world's problems.

concur

Marc wrote:...apart from the massive logistical problems involved...

slightly understated, but i guess i agree too.

Alright, with that settled, what about robots that play music! And paint pictures, write poetry etc. In what way will that compare to what we have previously considered 'Art' ?

...yeah, I'm just trying to egg people on, so?

Is it a sum of programming or a sum of experiences that donates their paintings the existential quality of 'art'

Glenjamin wrote: and if you're i'm poor, i can't help- this site opened my eyes a bit

If you put in that you make zero dollars a year it breaks the form and says you're the richest person in the world.

Also, I thought this would be of interest to you call:

http://en.wikipedia.org/wiki/Connectionism

It's about Connectionism, which is a broad term for a field of computational science that is involved with having multiple small processing units that each handle different types of computation, and connecting all of them to each other, MUCH LIKE A HUMAN BRAIN.

The most well known area of Connectionism would be Neural Networking, which is basically what I described above. It's a way of creating a computer that functions like a human. It has receptive units, that give the machine input (functioning like eyes and ears), these receptors process what they see and hand it to another unit that decides which part of the network the data should go. It's then sent there to be processed and output (unless it's processed and sent to another unit before output). All of this information is retained by the machine so it knows how to respond if it is given the same information from it's receptors, and can process it faster next time (Effectively, the machine is 'learning', based on a type of fuzzy logic).

Generally, the machine is programmed with preset instructions on how to process some information. For example, if it hears the word 'hello' it could be instructed that the proper way to process this should end with the output of saying 'hello' back. The issue with this is that computers are precise by nature. So if, for example, it was programmed to say hello back to a man, it may not respond in the right way to a woman (different voice frequency, etc). You would have to give it thousands of different scenarios so that it would build it's fuzzy logic up to the point where it would say hello back to everyone. Even after all that, it would be thrown by a person walking up to it and saying 'Hey'. Computers lack the 'close enough' way of understanding things. But still, with enough fuzzy logic the possibility is there. It would just take a shitload of data to be able to comprehend things in the way a human does.

See now that, to me, would still just be a computer with a lot of programming.

Like an Xbox?

Like a living, learning Xbox, yes.