[FuzzieDice]

You guys made some interesting points!

Dan - I see what you're saying. And your question is a good one. What level of intelligence? It might have to depend on the application to which the AI was created for, I would think. Some things, like answering the phone and forwarding to the proper rep if the answer isn't already in the database to be read back to the caller, that probably wouldn't need to be a very intelligent system. We have those in use now in some areas. However, one that has to say, calculate ballistics, and interact with military personnel, or one that works in law inforcement doing forensics analysis and profiling, and interact with investigators, that AI may need to be quite a bit more intelligent, as it were, in order to analyze and interact appropriately and accurately.

[Freddy]

Sorry, didn't want to interupt the flow but I think we uncovered how the language plays an important part in the understanding though.

Like my use of 'smart' confusing the difference between human smart and what a machine can do.  Without recognising it as artificial then thoughts seem to go off the map for me.

Reading back a few posts...If you want to know if you (we) can think Fuzzie, try asking yourself what you were doing when you replied to the posts   :wink  I'd say you were thinking.

A machine would be programmed to scan the posts looking for matches, patterns, inferences or whatever and send back some response.  So the real thought in those processes comes from the programmer I would say, the machine being a tool.

Coming back to meet you now ... if the above is true enough could/would the ai be as 'intelligent' as the programmer?

If Skinner wasn't worried about if a machine thinks, why was he more worried about if humanity does ?

[FuzzieDice]

Here's the clincher: While the programmer may have programmed the AI as it were, the AI should also have been programmed to LEARN from it's interactions and experiences.

Thus, you have two paths now. The programmer will go about their life and have different experiences from the AI who will have different experiences from the programmer.

Thus they could very well develop two different way of looking at things (ie. pattern matching, etc.)

Suppose the AI's video sees the grass is blue, matching it with what in the database is labeled blue hue. Maybe due to the fact the AI's video and color system is a bit off.

The programmer knows the grass is green, recalling being told that shade of caller is named "Green".

Now a colorblind friend of the programmer can say that the grass is a medium to dark grey. Because they can't see color, then they only know the shades of grey (they may have been told it was "green" but to them green can also look nearly or exactly like another hue of the same luminosity and saturation).

So, you have 3 different views of what the color of grass is. Is the machine wrong because it can't "think"? If that's so, then the colorblind person can't think which is why the person was wrong?

What I'm saying is, WE use pattern matching, which is "this is an apple" when we are a kid. "This is a dog. This is the color red." We program computers the same way because that is the only way WE know how to learn.

Things like this can get complicated to think about. Perhaps we should just THINK and not try to put some concepts into words because there is no way to describe some things. Thought, smartness, intelligence, life. Intangeable things that are up to a person's perception or view and while some have attempted to "prove" the existance of, they may not be able to CONVINCE someone ELSE of such findings.

Some may not think I'm actually thinking. Some may vehemently say I'm "ranting" which takes no thought whatsoever, and what I come up with is insignificant and of no real use (to them) and so they'll deduce that it's no real use to ANYONE.

Again, perceptions.

And I thank you for percieving my posts as "thinking". Thing is, I think too much and too deeply about things, often over-analyzing and I seem to either confuse people, or come up with descriptions that others can't fathom or rather not believe.

Communication is the #1 key it seems. But I think not HOW to communicate, but WHETHER or not to communicate. As they say: A picture is worth a thousand words. Maybe a thought is worth much more?

[dan]

Skinner was a behaviorist who raised his daughter in a box because he thought it was the right thing to do.  I'm not so sure "he" was worried, but the cause and effect of the statement and the controversy arising from the known behavior of the day being the reason for the statement.  Remember back then men were macho, women were barefoot and preg., etc. etc., and yet still see the controversy today, it shows how much we really are products of our environment.  I'm just a salivating dog at heart.  Now create AI with both behaviorist and inate (DNA), and yet the ability to know and advance beyond, with knowing it knows.  Then maybe someone might be fooled enough to believe it was sentient.  All that "who is the one that is doing the knowing" is enough to know it's hard to discern whether something is sentient.  I stink; therefore, I need a shower?   
Isdummy () { n++; };
cout << Isdummy::n << end;

I guess that's where "Fuzzy" logic comes in (no offense dice)...being able to C++ around enough to seem as if it is apparent to fool someone.  I heard the Haptek movement used a fuzzy logic, which makes is "seem" pretty realistic, although still looks fake, but more real than say an MS Agent.

[FuzzieDice]

Ok, I have to keep my mind on AI and chatbots here. Something hard to do when I'm always thinking of my project (which is none of the two). LOL!

I agree that one can emulate behaviors using certain programming algorythms. Thing is, what about the fact that computers can calculate and process information at a much faster rate than humans can? Would eventually it process things to the point of self-realization or will it be like the Deep Blue vs. Kasperov thing?

I think it all depends on the programming. The code itself is what it boils down to. Of course, the hardware (components that make up the machine) may have something to do with it as well. Faster computer = faster processing time leaving more time to process other things. Larger storage = more room to store and retain. And computers have a better retention rate than humans as well.

The hardware is there, now what about the software? Currently we don't have anything quite that good as to "think" as far as chatbots and AI is concerned (as I'm finding if I think strictly on those topics alone, and not other topics in with it). One I keep things separate, I see that humans can not "program" life - just yet.

But I think it's in there somewhere. I think that a machine can "think" if it had the right software. But, the programmer has to be able to think in order to make the program for the computer to be able to do the same.

Hmmm... Interesting to "think" about. LOL!