68 replies to this topic

[robocop]

Every human process can be replicated in a robot.....logically thinking. Imagine would a robot ever feel?

[anubis]

Quote

Every human process can be replicated in a robot.....

i doubt it...
seriously, it's terrible that the image humans have of themselves is more and more that of a (highly complex) machine. i'm not saying that i know it's wrong but it's far from being a fact

[JSoftware]

take humor.. we don't know even what it is! i can't realize that we can make robots understand humor when we don't understand it ourself

[Noor]

I, Robot :)

[anubis]

imo you can describe many aspects of humans in a way that might be reproduced by machines. but i fail to see how you could ever build a machine that, for example, has self consciousness

[EvilSmile]

Reminds me of `Cal' (a short story by Asimov). Please do read it if you could.

[robocop]

Let me try and explain what i mean. Talking about feeling, if u analyse what they actually are , in terms of the processes taking place in the body,we could relate them to some reactions or movements of the neurons in the brain.
In terms of technology(when it does reach that state), replicating these movements might someday be possible.Practically speaking it seems really tough.But it could anyday happen! :rolleyes:

[Mihail121]

Of course they can! It's just matter of time and progress. And at my opinion you don't need to create complex neutron network devices since a slightly modified human brain should also do the job fine. You just have to connect it and voila! Of course it's easier said than done but one never knows what will happen in the future!

P.S.
Wait if 'something' is using a human brain, then it's not completely a robot, right ?

[anubis]

Quote

Wait if 'something' is using a human brain, then it's not completely a robot, right ?

who knows... we know little about how we work and there are other centers like the solarplexus in our body that have a great influence on how we feel and behave (the saying that you can "feel it in your guts" doesn't come from nowhere). "connecting the brain", as you put it, might be a very tough job since the brain will expect all the input from the body and it's questionable that you will be able to simulate that.

Quote

Of course they can! It's just matter of time and progress.

let me reiterate : we don't know shit about how we work. even assuming that we can understand ourselves has no basis today. besides, this assumption conflicts with any spiritual concept we ever developed.

read "computer power and human reason" by joseph weizenbaum (MIT legend, the guy who wrote ELIZA)

[robocop]

Quote

let me reiterate : we don't know shit about how we work. even assuming that we can understand ourselves has no basis today. besides, this assumption conflicts with any spiritual concept we ever developed.

Its not true that we dont even know how we work.Talking about spirituality, we come across these ideas as we grow up, learning from the surroundings. When we say that we r trying to create or rather imagine a robot that would kind of be a replica of a human being first thing we gotto to take in mind is that it would be able to learn from the surroundings. This is also how one robot could be different than the other.

Mihail121:

The idea of connecting to the brain seems quite tough, cause hear we are considering the separation of a particular part from a whole interdependent system. Talking purely in terms of 'making' every part of the robot, the study and the working of neurons in the brain seems inevitable.

Anubis:
The feeling 'from the gut', will also be a kind of activity of neurons or say secretion of somekind of hormones/juices. It would boil down to a single process. I dont see why that couldnt be replicated

[anubis]

Quote

The feeling 'from the gut', will also be a kind of activity of neurons

that was much more of a statement about the interdependency between the body and the brain...

Quote

Its not true that we dont even know how we work

right, we know some things but in general we still know very little about how the brain works at all. part of the problem is that it's hard to conduct experiments on a living brain. so most of the experiments that are done reduce themselves to finding out which regions of the brain get used during certain actions or sensations.
a good example are these dumb devices we construct were you can concentrate on a left field and a cursor will move to the right while if you concentrate on the right field it will move to the right. we know that if a person thinks "left" hard ebough some region in the brain gets activated. fine. but how useful is a machine that requires me to think in a way that the machine can understand ? we don't understand at all what happens in the brain when we think "left" except for the fact that some electro magnetic pattern arises.

Quote

The feeling 'from the gut', will also be a kind of activity of neurons or say secretion of somekind of hormones/juices.

to get back to this sentence... just because chemicals can induce emotions that doesn't mean at all that emotions are fully described by mere interaction of these chemicals. for example : if we would for one second assume that there is a soul stuck in us it makes sense to assume that it is directly affected by the processes in our body.

personally i'm offended by the modern way of science to exclude everything from reality that we can't express in formal laws that our brains can understand and i would be more than amused to see that it's ourselves we can't press into that narrow scientific view upon the world

[robocop]

Quote

but how useful is a machine that requires me to think in a way that the machine can understand ?

i m sorry, i didnt get what u meant

Not knowing the working of the brain might be true if thought of literally.........i hope u r getting what i mean.See when talk of a robot that would be at par with a human being, i m not talking of us humans programming everything into the robot. What we r to do is make a robot capable to learn, for that we require the basic process of things going on inside the brain and the study of processes that trigger various reactions. That is definately possible.......to an extent i agree with u that as studyng the brain is tough........not impossible ( that might very efficiently happen whrn bio tech or bio med eng advances to that level).....so it might be easier to study the roots than the actual process .

Quote

i would be more than amused to see that it's ourselves we can't press into that narrow scientific view upon the world

what do u mean?

Also when i m talikng of robots at par with human beings , it is about being parallel and not the same .humans will never be equal to robots or vice versa , it is about 'living' parallel lives with interactions due to the intersection of the surroundings in each case.

[anubis]

Quote

Every human process can be replicated in a robot.....

Quote

Also when i m talking of robots at par with human beings , it is about being parallel and not the same

contradiction ?
do you want robots that mimic human behaviour or do you want robots that actually feel and think and could be thought of as equal to human beings ?

Quote

What we r to do is make a robot capable to learn, for that we require the basic process of things going on inside the brain and the study of processes that trigger various reactions

it took evolution billions of years to produce todays animals including us. so where do you start ? with the complexity of the brain a baby has ? a dog ?

[robocop]

Let me clear myself.
I m talking of robots that THINK AND LEARN BY THEMSELVES which can be considered equal to human beings.

Evolution has taken place for yrs together. We start with a human being only...... not necessary a baby...... all we need to know how the process of learning and thinking happens. The way, as of today we have been able to create light , sound or touch sensitive robots; in the same manner we could create a robot that picks up from the surrounding .ie using all senses and learn from it.

[anubis]

true. but a a baby can only learn to talk, walk, go to the toilet or eat with a fork and a knive because it already has a highly complex brain that allows it to learn these things. we are born with the ability to learn a language. we are born with the ability to walk some day. it's our nature because we have the needed structures in our brain (which is why a dog will never talk and a chimp will never solve differential equations).

creating a machine that has these learning abilites might probably be not much easier than rebuilding the brain of an adult

btw, what you are proposing isn't at all new. MIT and other universities have such machines... or had them. these projects had their time back in the 70's when they AI hype was big. today AI development is more about designing autonomous agents (like a robot moving through a room and cleaning it while avoiding all obstacles or wheelchairs that help disabled people to avoid obstacles)

[robocop]

Quote

btw, what you are proposing isn't at all new

I m aware of it.....that what i m proposing ia not altogether new. But as long as i know they have never had robots which could feel things emotionwise or robots that could be human like, they are working on similar concepts now and there are proper scientists also debating as to whether a robot could ever feel.

As of autonomous agents, i would consider it to be a step towards creating a HUMANLIKE robot; things like avoiding obstacles are the part of the processes a human brain does. May be we could call it as, what the research now a days is going on about or the designing taking place are very small fragments what a a humanlike robot would be.

I dont know if u r eamiliar with hardcore problems in math eg:combinatorics in those kinds of problems we bvreak it down into parts and proceed and at times what happens is that it becomes extremely tough to relate those minute parts of the solved problem.........probably in the near future we might head for this kind of a problem with robots!

Building a machine having learning capacities of a brainmight not be any easy.........may be .......but again as i was saying, replicating the various mechanisms rather than the ACTUAL process eg:neuron movements may be to an extent easier. To clarify myself; in say org chem we study various mechanisms rather than the actual attacking of the given compound on the other - simplifying it.

When we are born we are not meant to learn a language in particular, its just that we have the ability to pick up those things cause thats whats there in our surroundings.

[anubis]

Quote

When we are born we are not meant to learn a language in particular

no. we aren't born with the ability to walk either. and if you teach a baby to walk like a dog it probably will. none the less a child wants to walk at some point in order to explore the world. the same thing goes for speach. it doesn't matter if you teach a child a language or if you teach it to wave with it's feet and hands but it is born with the desire and ability to comunicate. on the other hand you won't find your kid suddenly communicating by telepathy, simply because we are not born with the ability to do that.

[davepermen]

actually, anubis, we know quite good how humans work. the ONLY issue we have is complexity, means, we're not capable of doing the whole work of a human body in realtime (but we can't simulate anything 100% in realtime.. we can only approx).

human life, animal life, plant life. feelings, movement, logic, behaviour, is all very controllable and understandable.

oh, and, it helps to know how feelings work. makes you.. feel bether.. if you feel bad. sort of :D

nature isn't difficult, nor magic. it's very simple, and easy. thats the beauty of it. the very simple rules that exist created the most complex thing any human being, any being at all, can ever be part of. a very impressive thing.

oh, and, about being on par.

i know tons of robots that can beat me, it just depends on what. a robot that 1:1 replaces humans is a dump robot imho. but we have tons of robots today, that are just here to be bether than us, a.k.a. to help us where we can't get a job done nicely.

i don't believe much in human-replacers. more in enhancers, that team up with us.

[anubis]

Quote

a robot that 1:1 replaces humans is a dumb robot imho

yeah, who would want a robot complaining about his wages or demanding better social security :)

Quote

actually, anubis, we know quite good how humans work.

afaik, they have problems understanding the brains of flys...

Quote

the ONLY issue we have is complexity, means, we're not capable of doing the whole work of a human body in realtime

you sound like every part of the brain is SO well understood and that it's just a matter of reaching a certain complexity in our simulation capabilites. bullpies to that !

Quote

human life, animal life, plant life. feelings, movement, logic, behaviour, is all very controllable and understandable

i'd like to see your proof for that. just because it seems simple to you doesn't mean it is

[davepermen]

all structures are well understood and simple. it's just, they all work together, and influence each other. and this, by today, is, even in case of a fly, out of our control.

thats why we can't understand a fly-brain. not because it's something strange, magic. the rules, with wich it works, we know, we can identify each component exactly, and what the part does. but we can NOT understand, and work with the whole brain at once. we can't get up to that complexity, it's much over our head, and computing-simulation-power.

it's similar to chess, understanding the rules is easy. winning against a master is near to impossible. you know the rules. you _should_ know how to win. why don't you? ... same reason we can't really use the knowledge we yet have. thinking on the whole isn't doable. but we can correctly identify each component, and what it does, on it's own.

proof? you know drugs yourself... :D there's tons of proof out there. medicine, sports, all the stuff, that has to do with our body. it's all just about how it reacts on different situations, influences.


we know the human brain about as good as you, and i can program c/c++. quite good, we know about all we need to, and can read all sort of code. the same way, we can read info in a brain today, easily. we can see the "brain-code" of your fly.

but even while we can read source-code, i'm not capable of reading the whole windows XP source, or the full linux source, and simply "go in there, and fix some problems".

and THATS nature. a human body is not billions of lines of code to read. it's sorta billion billions. while we can understand it's behaviour, we can not control that mass at once. not anytime soon.



and i still don't want robots to take our job on being "the dump-ass of the world" :D