Theoretical question about AI

Post Reply New Topic

It is all possible,
all you need is money, time and some brilliant minds, as myself and some of you who read it...

I don't know about pain. There would have to be special pain sensors installed on the robot (not the AI brain) since pain involves signals being sent back and forth from the body to the brain. But emotions, sure. An AI brain could be programmed to have emotions, or if it were sufficiently complex, maybe it could even develop emotions on its own (highly unlikely, but possible).

That's already been done using simpler technologies.

Fully operational AIs in autonomous form would be more convincing to most people, though at that point I suppose the "machines" would have to prove their not homo sapiens, and homo sapiens may have to prove they are to machines.

Why would you give it emotions?

It would wind up having a breakdown and end up in therapy.

it is called wet-ware

and if you build an ai that responds to chemicals, it isn't unlikely that you could stimulate pain or pleasure.

emotions are basic motivating factors. no emotions would leave you back at basic computer. Even basic emotions such as avoiding pain, desiring to reproduce and remain alive and seeking pleasure are necessary for all life

I already have AI which can be in the form of a monkey that writes Shakespeare on it's
first try, or makes any music, but it only generates it as binary data that goes to a
printer or a speaker, and doesn't know or care what it is doing, it just knows all
possible sequences of binary data instantly and is therefore just an infinitely fast
binary counting machine. Anything and everything inside a computer can be invented
by a simple binary counter that spends all eternity counting and trying each binary number
as a file to see what it means. Every possible file is just one binary number.

Now, I've recently discovered, or at least it seems to me so, that anyone watching TV
only absorbs about 64K per minute maximum, and that it's possible to computer generate
a minute of watchable TV using LESS THAN 64000 BYTES. (If a picture is worth a thousand
words, then try even typing a 64000 letter essay on what you learned from watching one
minute of TV, for example, as a semi-proof against my theory.)
The results imply that we don't see all we look at, and our computers are already smarter
than we are; perhaps we are too dumb to program them to be like us.

But I think that chemicals are required to feel sensations the way we do, at least on the
philosophical argument that an electromechanical android copy of us cannot get drunk.

To create an ai that replicated humanlike thought processes, it would likely require that it must be a learning ai. It's initial data set would only include "instictual" information. And some form of motivational factors would need to be implemented to get it to do...anything.

One of the key features of our minds that I personally believe contributed to self awareness, is the degree of interconectivity of various specialty regions of our brains. Where each section of our brains has a function, and a purpose. And at times the regions cooperate, and sometimes they conflict. There is conflicting emotions, thoughts, and motivators and they are all interconnected. The results are, imo, obvious in our behaviour as a species.

Additionally, another aspect to noticeable self conciuosness, and humanlike behavior. Is a data filtering system. We take in far more information that we are capable of recording for practical use. As such, we filter that data into what is relevant and dump the rest. With out something similar an ai would be too preoccupied with gather data and information to even function.

>.> Sounds like something else I know of.....