self-awareness and the desire to be known

[asakiyume]

I've been musing on self-awareness and on humans' desire to have machines be self-aware ever since the story about the guy at Google came out. My thoughts have run in all kinds of directions. For instance: about relationships up and down the awareness scale. Most of us likely have had relationships with beings more self-aware than we are (parents are generally more self-aware than toddlers, and all of us have been toddlers and had parents or others filling that role), and most of us likely have relationships with beings that are less self-aware, and/or differently aware, than we are.

Those relationships are not only with living things but with nonliving things: we have feelings about and express ourselves to our computers, phones, cars, coffee makers, microwaves ... These might not seem like relationships because they're so one-sided, but I think they are: we interact; they respond to our inputs; we respond to theirs. We don't expect our microwave to discourse with us on anything, but we do expect that if we press a button, it will shoot microwaves through something and heat it/cook it for us. We're happy when it meets our expectations and disappointed or worried or annoyed if it doesn't.

What I'm trying to suggest is that we have relationships with all kinds of things of different levels of awareness, and we're generally fine with that. But the more like us something or someone is, the more we seem to want its/their awareness to match ours. Misunderstandings that arise with people very close to us show how much we expect or depend on those close ones' awareness matching ours. But sometimes it doesn't. We say a thing, and to us it's pregnant with meaning and import, and the person we're talking to replies, and we feel they've understood! Their thoughts are running the same way, and their response shows that! Only to discover later that no, they were *not* thinking in the way we imagined, and furthermore, they had no idea that what we said carried so much weight for us.

Or we can be on the other side of that--having an innocent conversation one day, only to find out to our alarm that it had all kinds of other meanings for the other person.

Those differences are painful, but it would be a weird kind of tyranny, a kind of Borg-ness, to expect another human being to understand and respond to us perfectly ... impossible really, given that we can't even say, ourselves, what a perfect understanding or response would look like.

I was thinking, if a machine/AI could be so cleverly programmed that it could duplicate human-type reactions, human-type non sequiturs, human-type self-absorption from time to time, but also human-type friendly queries, supportive remarks, gratifying curiosity and so on---all based on code--would it matter that it was code that was generating those responses and not whatever it is that generates those things in a human? Could being in relationship with a machine/AI on its own terms mean accepting its machine-ness and not requiring it to duplicate organic human-ness?

What do you think?

Comments

[cmcmck]

I sometimes wish my camera would give me a prod when it notices something I've missed even though I'm pretty observant over potential shots! :o)

[asakiyume]

I feel like cameras that can do this are right around the corner ;-)

[sartorias]

Misinterpreting, or perc living things from a radically different perspective, is certainly a big part of the human experience. I wonder if AAIs, given enough time, would perceive all alike, or if they, too, would alter according to experience.

[asakiyume]

I sometimes imagine a machine/AI society, which creates offspring according to its needs/desires, and when I imagine it, I wonder if the programmer-machines, creating offspring machines, would introduce some kind of randomness factor, kind of the way biological organisms have with reproduction...

[amaebi]

They might have no choice about the randomness.

[asakiyume]

That's possible!

[minoanmiss]

but it would be a weird kind of tyranny, a kind of Borg-ness, to expect another human being to understand and respond to us perfectly ... impossible really, given that we can't even say, ourselves, what a perfect understanding or response would look like.

... this explains a lot about my childhood.

I think you have some really excellent points here about human nature and the nature of consciousness, but I admit this made me think about disconnects with my parents, with the kind of consciousness involved in being "Christ-minded" vs not.

[asakiyume]

The disconnect with your parents to me seems completely natural (and very painful/hurtful)--that's them being Borglike with another human being: you. They wanted to--no, not wanted to, insisted on (despite evidence to the contrary) your conformity to their vision. But that's not how you were made.

... I'm sorry it [by which I mean, this post] brings up painful memories, though.

[minoanmiss]

Oh goodness no apologies necessary! husg you

more

[asakiyume]

I think one thing I wanted to say in this post, but I'm not sure I actually did say it, is that we can only interact with things as they are, for who they are--not by wishing or asserting or imagining that they are something different.

And we have to acknowledge that other beings/things/people can only interact to us and respond to us according to their natures--which aren't static, and which we may not understand, so it's not like we should give a thing or a person a label once and for all and decide that's all we can expect--I'm not trying to say that--I'm just trying to say that we can love the clouds for being clouds without expecting them to interact with us the way our best friend from childhood used to interact with us.

And I feel like machines deserve that respect too: let us appreciate things that run by their programming or by their circuitry and wiring, how well they do what they are designed to do, without always yearning for them to be something else. Not that they'll mind if we do yearn, but it's probably better for our interior landscape if we're not always yearning for something to be something that it's not.

--When I look at that written out, it sounds rather negative, like a person saying, "Don't expect a woman ever to be more than a bobble-headed idiot; that's all she's capable of"--but I don't mean that. I only mean that we should take things/people as they are--as best we understand them--and let them show us when/if they're more/different. (But if they do show us, we need to be open to that and accept it.)

Re: more

[minoanmiss]

What I see you talking about here is making sure that our perceptions align with reality (which can be a lot harder to do than people expect -- all our perceptions are filtered through our senses, our interpretations, our history and knowledge or lack thereof...) and I think that makes a lot of sense.

Re: more

[asakiyume]

((hugs)) I'm glad

Re: more

[amaebi]

You communicated it to me. To my delight.

Re: more

[asakiyume]

This makes me very happy!

[sovay]

Could being in relationship with a machine/AI on its own terms mean accepting its machine-ness and not requiring it to duplicate organic human-ness?

I would like to think that we would have to, if it was truly sentient and we wanted a real relationship with it, just as we have to accept that other people are other people.

[asakiyume]

Me too--and even if it isn't, though if it isn't, it won't "mind" the way we would mind.

[sovay]

though if it isn't, it won't "mind" the way we would mind.

And you don't have the same kind of relationship with something that isn't sentient as you do with something that is.

[asakiyume]

I think I agree with you, but I think the kind of relationship we have is more based on our understanding and interpretation of the aite (*such* a convenient Japanese word that I wish we had a direct equivalent for--it means "person you're interacting with," could be interlocutor or partner or competitor or whatever) than on anything inherent in them. In other words, if we think we're interacting with a sentient being, that changes the nature of the relationship in our mind--but we may or may not be correct in our assessment, and our sense of the relationship changes based on our impression, not on anything in the aite. .... which I guess is true in any relationship too? If we think someone's in love with us, it changes how we think about them and react to them, but we may be wrong in our assessment--but it's not their feelings about us that determine how we interact with them but our understanding of their feelings. ... I feel like maybe I'm belaboring an obvious point, and if so, I apologize. It's one of those times where I'm not sure if I'm saying something glaringly obvious that other people don't even bother saying or if I'm saying something worth saying.

[amaebi]

I never know how self-awareness can be well assessed from outside, nor one's own compared with that of any other entity. Though we humans do a lot of it.

I have never understood offhand dismissals of animism-- not that you're dismissing anything! How on earth could one know about the consciousness of animals, let alone plants, fungi, waters, stones, the earth and moon and stars...? And electronics, which like us operate on electrical rates, have always seemed to me to bring their own sort of kinship with them.

I love what you say about relating to AI as itself, not as faulty-us.

And how I wish humans more usually treated other humans that way.

[asakiyume]

We have to be doing so many things at once: both applying the Golden Rule, but also understanding that other people (or beings that aren't people) may NOT have the same wishes and desires as us, so in some cases we should NOT apply the Golden Rule; also, understanding what people/beings are (trying to, anyway) and not insisting they're something else, but at the same time being open to a change in perspective on what they are. We have to always be adjusting--like keeping balance on board a ship.

[wayfaringwordhack]

Interesting thoughts. I have automatons in the book I am working on, but not on a fantasy/sci-fi level of being sentient (though in the next book there will be an advancement in the tech). Automatons are far from the point of the book, but I listened to an interesting radio debate on humans' desire to create automata and conversely the fear of said creations and a look at that relationship through history.

[asakiyume]

I think humans have human-centrism (understandable--humans are what we are) when it comes to interacting with other things or thinking about other things. So we want intelligent machines to be like us--because "us" is the best way to be!<--that's the human-centrism at work. But we are afraid because (a) we also know that humans can be pretty awful, and the more us-like machines become, the more likely they become like us in the awful ways, too, and (b) the unknown. What if intelligent machines are different from us in ways we don't understand? That scares us.

(By us I don't necessarily mean you or me, personally, or all people, but people in general.)

[wayfaringwordhack]

Yep.

And the "scary" factor makes me think about our other convo about using insects as fodder for creating monsters. We have a tendency to look at the worst or scary parts of something and amplify it enormously.