Now you are backpaddeling . Behavioural AI is what we got today. You define the behaviour as a programmer and when to perform that behaviour and machine performs. What you wrote is implementing piece of code into a living organism and it emerges its own and unique decision from sub programs but that's just a science fiction for us now. In fact, we are going that way but we are not there yet. There are researches on modelling AI based on biological structures, not structure itself but rather and abstract respresentation of let's say cell and each structure has its own duty which allows evolution. That's called biological-inspired programming(mind the word "programming", not implementation into living organisms). Also, note that neural networks(which how our brain works) proven to be bad problem solvers and our algorithms are not advanced enough to simulate how biological structures work. I was not incorrect, you just overestimate the state of AI researches. Robots will be stupid for a loooong time, again.
Last edited by Kuntantee; 2013-02-24 at 12:12 PM.
Nice try at mangling burden of proof.
If I want my assertions to be accepted then yes, I hold burden of proof to back them up. However; that burden of proof is met by advising where people can go to look for information. In this case, google and a pair of search terms.
Burden of proof does not mean I have to hold someone's hand and lead them straight to a relevant source; it means I have to have a source on my side.
Intellectual laziness coupled with snarkery is not likely to get anybody to want to make things easy for you.
Having just read the Metamorphisis of the Prime Intellect, a scary thought. But an interesting one nonetheless. Humans at their base are little more than complicated biological machines, so it is in no way inconceivable that it is possible to create an artificial machine that can become self aware and as 'alive' as anything else.
I think a better question is, can a machine learn to be creative and come to have its own opinions.
Last edited by Herecius; 2013-02-24 at 12:12 PM.
Programmer my self. Absolutely not. It will and only will do as instructed to do so. In fact it can be frustrating because codes often do EXACTLY what has been instructed, so any slight mistake (eg. missing a ";") in your miles of code can cause it to behave in unexpected ways (to a human that is).
only if we program them.
I cant wait until i have a computer with a smart AI then playing http://www.youtube.com/watch?v=h03CshUyvCc will make it even more creepy then it already is.
ROFL! Please tell me you just completed Philosophy 1000. I am one semester away from completing my PhD and here in the read world we have whats called burden of evidence where yes when you make statements such as yours you better have credible evidence to back it up. If you trying making bullshit statements such as yours calling people lazy for not looking it up themselves you will get laughed out of any room.
And if you try and whip out some more theories you just read about I am walking away from this one.
Argument from authority is a logical fallacy. Also, I find it amusing how frequently when someone is challenged, they're suddenly an expert in the field, and 'one semester away from my D.Phil! Honest I am!'
Also; it's nice how you're quote-mining, if you are about to get your D.Phil you really should know better than that by now.
And finally; my statements were not bullshit, as anybody who'd take the very simple steps to locate support for them would have found out.
The only way you're about to get a D.Phil is from a back-alley, one room, 'university.'
Wouldn't such a system, if asked the same question at the same point in its development each time, always come up with the exact same answer, though? That's more what I was talking about. The ability for spontaneous creativity, linking abstract, oftentimes unrelated concepts together to form opinions as a human being does.
Again, wrong. It is exceedingly trivial to create an ANN that can, for instance, classify the Iris flower data set. This may be a simplistic problem, but it is problem solving nonetheless ("which type of iris plant does this belong to?").
---------- Post added 2013-02-24 at 12:26 PM ----------
Only if there is no random element involved in the training process. A randomised initialisation of the starting weights solves this.
1) Not related at all to what you just said about "at the same point".That's more what I was talking about. The ability for spontaneous creativity, linking abstract, oftentimes unrelated concepts
2) This would be a very advanced AI, but there is no reason to assume that this can't be achieved.
Yup, using big boy words such as logical fallacy, you truly are the armchair debater.
Your statement was a scientific one, not a philosophical one so your "burden of proof" is complete crap.
And in regards to my background, ask me anything you like regarding AMP-activated protein kinase and ACC and their role in the modulation of cellular energy states following myocardial infarctions.
Still waiting on those references champ, or are you going to back pedal even more and use words like "strawman" in your next post?
not "self-aware" enough to be mistaken for a human or a real "AI"(thinking for itself, doing things for itself, living for itself) but it's getting pretty close
it's only a matter of time really, it would go by faster but alot of people put severe restrictions on research of this kind (i.e funding etc.)
Ahh, ye olde ad hominem, coupled with a complete inability to assess the matter at hand.
My statement was not a scientific one, it was an information-sharing one. A scientific statement would have been an explanation of the fundamentals of behavioural AI, and not a 'hey, this is a thing! It's neat! Check it out!'
And again, one which is born out by actually taking the time to plug the search terms in to google.
In regards to your background; not only does it not matter what you can copy+paste off the internets, it also doesn't matter whether or not you actually do know that. It's a rather shoddy attempted argument from authority.
The references, or rather the procedure needed to find them, has already been posted. It's kind of hard to 'back pedal' when the destination has already been reached!