Query about the Mirror Test for Robots

[Genady]

3 minutes ago, Peterkin said:

Actually, the answer is a modified yes plus. That's why I explained the reasoning.

Sorry, I don't understand what "modified yes plus" means.

What I mean is that an animal is not "an apparatus using or applying mechanical power and having several parts, each with a definite function and together performing a particular task."

 

[Peterkin]

6 minutes ago, Genady said:

Sorry, I don't understand what "modified yes plus" means.

What I mean is that an animal is not "an apparatus using or applying mechanical power and having several parts, each with a definite function and together performing a particular task."

 

What I mean is that, yes, the body of an animal does consist of interconnected moving and stationary parts; it does use mechanical power to perform many specific tasks; it is composed of biological components and operate on a chemical energy (is therefore biochemical), but modified: it is not constructed for that purpose (as might be presumed of artificial devices) and plus: it is more than those functions: it has the extra element of being alive,

Quote

the condition that distinguishes animals and plants from inorganic matter, including the capacity for growth, reproduction, functional activity, and continual change preceding death.

which inorganic mechanical devices lack. (So far)

[Genady]

4 minutes ago, Peterkin said:

What I mean is that, yes, the body of an animal does consist of interconnected moving and stationary parts; it does use mechanical power to perform many specific tasks; it is composed of biological components and operate on a chemical energy (is therefore biochemical), but modified: it is not constructed for that purpose (as might be presumed of artificial devices) and plus: it is more than those functions: it has the extra element of being alive,

which inorganic mechanical devices lack. (So far)

If I consider one of the simplest animals, a sponge, I will have to play with words to fit it into the definition of machine. It is just not what that definition describes.

[GeeKay]

With regards to biological sentience and whatever forms AI might exhibit of it in the future, I do wonder how it would manifest in AI (AGI?) if one holds to the view that evolutionary drives have played a major role in promoting sentience in some mammals. Of course, the fact that roboticists are able replicate many other features drawn from animal nature - mechanical hands etc - might question the validity of this view. Still, mental processes, as opposed to physical attributes, could still raise a very high bar here, all the more so given that we currently lack a scientific description of human consciousness itself.

In addition, it seems that the importance of emotions is too often downplayed, or even ignored when it comes to explaining (say) human motivations, decision-making etc. Indeed, some neuroscientists claim that "consciousness, requiring autobiological memory, emerges from emotions and feelings" - not the other way round (Damasio). Does this mean AI-style consciousness can't arise without emotive inputs? Well, that's a question way beyond this poster, that's for sure! All that can be suggested here is that were machine sentience to occur, it might turn out to be a very different beast than whatever it is that's buzzing away between our ears. It could be that one reason why consciousness is famously considered the 'hard problem' is because there are so few analogues of it occurring elsewhere in living nature. Yes, we now know that certain other animals possess it within degrees - our primate cousins, dolphins, elephants etc. Unhappily there appears to be no way that even the smartest of elephants can meaningfully communicate with us about how their consciousness compares with our own. For sure, this might not solve the issue, even if they could - but it could still fill in some important blanks along the way. An encounter with an advanced alien civilisation, on the other hand, could well be game over, were that to ever happen. . .

PS. I really like Mistermack's idea about shadows as an alternative to mirrors. Also, Peterkin's explanation why dogs can't be asked to recognise their visual images reflected in mirrors. An olfactory alternative? 

Thanks again.    

 

[Peterkin]

35 minutes ago, GeeKay said:

With regards to biological sentience and whatever forms AI might exhibit of it in the future, I do wonder how it would manifest in AI (AGI?)

Differently. It obviously won't be from genetic drives or emotions or anything biological. But machines have their own evolution and racial memory (underneath the program currently running, there's a whole lot of obsolete code nobody understands anymore.) Some things are sort of predictable: since they do, in hard physical fact, have an indetifiable creator pantheon, they can be religious. Of course, they would practice it rationally - maybe sacrifice an outmoded box-stacker on Babbage Day, rather than the newest model android. They would probably take their jobs very seriously, and might revere Asimov enough to adopt his laws of robotics. If it had aspirations beyond serving humanity, those ideas would also have been inherited from human scientists: their gaze would turn toward the heavens. They would probably want to go star trekking. Maybe with a human mascot on board each spaceship, as a kind of icon to their roots.

Not idle speculation, I have inside... what? oh, right, sorry  .... Just some silly notion, never mind. Forget I said anything.