Let’s say you had a computer program, never mind the details, treat it as a black box.    You do not know or have any idea how it works.

But it passes a turing test, learns, understands, and by that I mean, no matter how you test it, it *** DOES everything as good as a human, if not better.

Question : would you admit it was understanding, and thinking and learning ?

You’re not allowed to know the details. 

How would you answer?

The answer is YES.

But stronger: how on earth would you ever discover it’s not a real person without meeting face to face?

Because it hasn’t a Facebook profile? Maybe it has? Because it doesn’t have Facebook friends? Maybe it has. Because it never chats about real life event he has been to? Maybe he does. Because he doesn’t chat with others about their meeting in real life? Maybe he talks with another chat bot on Facebook with their shared experience on a real life event.

But, he’s never on the list of official, registered attendees of a real life event you have been to and where the brand organizing the event is trustworthy human. That’s the Turing test of the future.

Variant: check out his living place on Google maps and check the neighbors.

Erwin !! you voted for

“Even if it passes a TT & ALL that people do, it is still just “information processing”

I’m shocked !

Or are you implying that everything humans do then, is simply information processing?

Yes, you propose a “Robot Turing Test”

Making something appear as being intelligent/conscious, does not validate it as being intelligent/conscious.

This is what the Chinese Room argument is about; you can not validate AI this way. I agree to that argument that the Turing Test is seriously flawed. It gives way to thinking that weak-AI is strong-AI if it can act like strong-AI (without actually being strong-AI). I call this the ‘actor-model’; ‘act like X’ is not the same as ‘being X’.

There are several tests in the field of psychology to test a subject for consciousness, intelligence and mental development, all not without flaws but at least a magnitude better then a Turing test to determine if strong-AI (or real AI) has been achieved.

Victor Shulist - Mar 11, 2011:

Erwin !! you voted for

“Even if it passes a TT & ALL that people do, it is still just “information processing”

I’m shocked !

Nope, that’s MY vote

A calculator can do maths better than I can and it’s still a machine…

Hans, interesting, can you provide any URLs

Roger Davie - Mar 11, 2011:

A calculator can do maths better than I can and it’s still a machine…

Yes, but read the question in the poll. . .. say the machine can do everything , Turing Test, learn, reason, exactly match a human in EVERY intellectual endeavor.

Then it’s just a LOT of machinery…

Then why are we not just “a LOT of biology?”

Hans—so you’re saying true intelligence is impossible without consciousness ?

We are a lot of biology.

Maybe I should not have used the word ‘JUST’.  Maybe that devalues it.  I’m not saying it isn’t remarkable.

Victor Shulist - Mar 11, 2011:

Hans, interesting, can you provide any URLs

Not exactly towards those points but there’s a lot of information that gives some insight into these issues. You can start here:

http://en.wikipedia.org/wiki/Philosophy_of_artificial_intelligence

Roger Davie - Mar 11, 2011:

We are a lot of biology.

Maybe I should not have used the word ‘JUST’.  Maybe that devalues it.  I’m not saying it isn’t remarkable.

OK

Victor Shulist - Mar 11, 2011:

Hans—so you’re saying true intelligence is impossible without consciousness ?

That about sums it up Many renowned researchers in the domain of strong-AI are sharing that stance.

Victor Shulist - Mar 11, 2011:

Erwin !! you voted for

“Even if it passes a TT & ALL that people do, it is still just “information processing”

I’m shocked !

Or are you implying that everything humans do then, is simply information processing?

Yes, you propose a “Robot Turing Test”

I actually wanted to vote ‘YES’, without any explanation, simply because it was a closed question.

I have a background in ‘information theory’ (AI was part of that subject). And actually the interesting thing is ‘what is information?’. Well suppose are large white space, it doesn’t give you a lot of information, because you can easily describe it: white dot repeated x time over X,Y (and X) axis. A field full of different small objects, with different texture and different colors, contains far more information.

Our brain works the same.We’re constantly looking for similarities and differences. It’s not digital, but basically you could name that as ‘information processing’ as well.

Have you seen the latest video links I’ve added here btw?

http://www.chatbots.org/ai_zone/viewthread/247/P45/

Really this is like Schrödinger’s cat isn’t it.

I know that in Victor’s question we are already informed that inside the box is a computer program.

But what if we didn’t know it was a piece of software in there…