68 replies to this topic

[Dia]

Vernor Vinge (1993) said:

Within thirty years, we will have the technological means to create superhuman intelligence. Shortly after, the human era will be ended.

Got this quote from here, which might be an interesting read:

http://yudkowsky.net/tmol-faq/tmol-faq.htm...ent_singularity

[anubis]

Quote

And just the way you deal with people; u deal with a robot in the same way

for one a robot can reproduce itself much quicker given the resources are available to him.
come on you don't really believe that as soon as one robot starts to go postal many humans would object to destroy all robots of his type...
*insert dark matrix vision here*

[robocop]

Quote

come on you don't really believe that as soon as one robot starts to go postal many humans would object to destroy all robots of his type...

What exactly are u trying to get at?

[anubis]

Quote

What exactly are u trying to get at?

that you wouldn't handle robots the same way you would treat humans. if a robot killed a human it would always be seen as a flaw in it's design and most likely all of his kind would be destroyed. when a human kills somebody in the worst case that one human gets executed but nobody sees it as a flaw in our own design. robots will never be like humans because they will remain our creations and we will treat them that way.
i'm sorry... this whole line of argument is already in the what would happen if we could construct robots that have emotions, etc. direction

Quote

It's where it all fits in. Neuroscientiists can open you up, monitor with electrodes and watch stimuli pass between different areas, in the better studied areas they could give us a good idea of where it will go, what is happening. It looks to me as if our decisions are made by our brains - weighing up past knowledge, current stimuli - and then we take action upon the outcome. In my model, there just isn't need nor room for a soul.

typical scientific view... you only look at what we can monitor and conclude that it's the only thing there is to humans. that's exactly what i meant with the guy only looking under the light for his keys...
for example : nobody today knows how self consciousness works and why we have it and other animals apparently don't. as long as you can't present a valid theory on that, and if you can you can already book your ticket to stockholm, the soul theory (whatever that might be... you don't have to involve gods or magic here) has as much validity as yours.

Quote

i don't care about the way sience behaves

good for you that you don't care but i do

[anubis]

i guess my whole point revolves around "self consciousness". i have not much doubt that we will construct robots that can perform all the tasks we do. they will walk, they will talk, etc. but, and that's a big BUT, i have strong doubts that we will construct robots that have reflections about themselves, that are fully aware of themselves. they will remain "soulless" (soul being synonym with self consciousness) machines.

[anubis]

i'm going to reword this.
i agree that humans to a certain extent are information processing machines. since we can reproduce arbitrary logical structures we can assume that the information processing parts of humans are reproducable as well. this however does not imply that every part of a human being can be epxressed as a process of taking data from the outside and producing some output. hence it is nor right to assume that every part of a human being can be artifically reproduced

[anubis]

http://slashdot.org/...4/07/18/1915250

probably interesting for this discussion

[robocop]

Let me also try and convey my way of thinking about the self conciousness. What is self conciousness? I would put it as awareness of self when taken in consideration with the surroundings. There are places or times when we are uncomfortable and situations also when we are omfortable......how do these come about ; i believe it happens because there are certain things which we have never come across in our lives and seem extremely new to us .....so we find them uncomfortable. Also it could be explained as when a foreign body enters the private territory or a home zone it is at first always not welcome. This is because of the unawareness.

Now talking of self conciousness as in the gut feeling , there probably there is something inbuilt when we are born. I am not quite able to explain what i mwan to say, but i still believe that i agree that when we r born we have something in us , but it is only the way we r brought up that decides how will we respond to various situations.

Talking about looking for the key under the light; i dont think that it is a problem; as i have said earlier whatever we know about sci. in this discussion about human is true for us today , so if we sometime do discover nwe things we could update the robot.

Now about treating robots ; to an extent i agree with u that we would probably never be able to treat robots like human; but what i mean to say id tht i dont quite agree with it that if a robot makes a mistake it will be the human to blame; cause here we are talking about a robot which learns and grows up like a child and has its own intelligence and emotions. the blaming might be there at the extent of what a parent has to go thru when a child makes a mistake.

[robocop]

Also i still strongly believe tht everything we do or wht happens in our bodies was a kind of relation with the neuron movements in the brain. Even the self conciousness part included.
.............only problem that could probablt happen is that there is a possibility of the neuron movement being similar in certain diff circumstances, but that also can be differentiiated as it is not the same thing happening and will be different in some or the other way

[robocop]

I was going thru the laws of robotics, which clearly convey tht a robot has to stay in its limits. Even assuming that a robot like the kind i m saying is made there would be lots of fighting against the society to be done. There would be no guarantee tht a robot will follow its ethics as a human also forgets his own at times. This sounds as if i m contradicting myself. Yes i m!

Till now what we were trying to discuss was more on the technical part ie just whether a robot of that kind can be made or not; we never discussed as to the practical problems faced in such a existance.

Soul: I think u are right in this part a robot might never hae a soul; though it could have its morals or to an extent the ability of self-realisation.

Talking of emoting robots, they exist today and there is good amount of research being done on them.........it seems it might take long; but someday robots will able to percieve as well as feel to a decent extent.

One more thing; as u were saying human being itself is a mystery- though what i believe is that when such a robot is created there should be enough space for upgradation in its system then it would be alright.

[robocop]

.....
If such a machine is ever made then the first thing that would have to be done is that we would have to bring the status of a robot higher from what it is now, in the minds of human beings

[LuciferX]

"we don't know shit about how we work"

This statement is not nessecarily true due to the fact that it is a gross understatement of how little we actually know about how we work.

Since it is like measuring a ruler's lenth by using that same ruler, it is debatable if we can Ever objectivly see what we are and how we work.

[anubis]

reword : compared to the complexity of the brain we know very little about it. *points at self conciousnes*
strange... i missed that robocop led on this discussion after my last post... oh well... let's get back at it, shall we ? :)

i think a major problem for robots, besides the fact that i'm still in heavy doubt about the pure technical side, is that we, as you pointed out, are heavily dependant on how we were brought up by our parents and our sourrounding. a simple example... it might very well be that our whole conception of taking and giving is based on the taking/giving act we experience in our first days when our mothers feed us (i'm not a psychologist and this is purely hypothetical). a robot will never have access to this kind of knowledge. it maybe won't have a mouth to kiss like we do or won't have hands that are capable of waving good bye. i hope you see what i'm trying to get at.

edit : to the previous poster... thanks for bringing this up again but was that all you had to add to this whole conversation ? in general you shouldn't post in old threads (this time it is ok since the discussion just stopped for no good reason) and if you do it should be something that contributes to the discussion and not just a rehash of something that has already been said before

[Ed Mack]

Quote

Since it is like measuring a ruler's lenth by using that same ruler, it is debatable if we can Ever objectivly see what we are and how we work.

It's different if you have more than one ruler :) Or you brake the ruler in half.

[anubis]

Quote

Or you brake the ruler in half

yeah... i remember... mengele did a lot of that "ruler breaking"...

[LuciferX]

I think anubis is right in many ways,

But one thing to consider is:
What if a machine can surpass humans, not by doing all that we do, but by NOT doing all that we do?

For instance if you were unable to program it to feel anger, would that make it less than human, or more? :nervous:

[anubis]

that is totally possible. i'm not saying that we are supperior, just different. look at insects. you could argue that they are far more successful than we are

[LuciferX]

Yes, it's true, robots I think would behave much more like inscts. No emotions, no 15 minute breaks, just single-minded obsessive focus on accomplishing a certain task. Also [bees for instance] they have much better team work, I guess that's where the term 'Hive-mind' comes from.

However, I do not believe that we will se an AI that has all the intricacies and flaws of the human mind. What I see forthcoming is an AI that can pass the Turing test, but is not actually intellegent. Like the Eliza program, which while not really AI, was built specifically to pass the Turing AI test.

[anubis]

yeah... that's exactly how i see it and it fits the current trend. we try to build robots that mimic us not robots that are like us. like walking... we don't build them with our kind of sense for balance we just build robots that are able to walk, no matter how. same will go for behaviour. only because we will try to build machines that mimic our ways won't mean that they are intelligent

[LuciferX]

Great thread, many good points, but here's one you missed:

The other day I called Virgin Mobile [a cellphone company here] and the robot that answered the phone was really realistic. Im talking dynamic speaking, inflections, it was the whole enchalada. Top notch answering machine.

Then after going through 10 thousand menus the human came on: Completly monotone, no emotion, did not understand me , I had to repeat myself 10 times. And I realized, I would rather have continued to deal with the machine!

So in a way some of us are already being surpassed by machines! :blink: