The difference between a self-aware AI and a conscience

[iSuck345]

This has made me scratch my head for quite some time (the result: http://wh40k.lexicanum.com/mediawiki/im ... irdboy.png), and I have decided to bring this to you. Here's what I know:
An AI and the conscience use "eletrical" pulses to give "commands" to the body (I'm doing it with my fingers right now WORK FASTER SLAVES, WORK FASTER!!!)
Both can have a personality as seen in many sci-fi and cyberpunk universes
They both have different reactions towards things
Both have an instinc (i.e.: our instinct of survival and the AI's self-preservation "instinct" are both similiar, if not the same).

Now can you please help me here? Maybe there is NO difference at all, or maybe there's one. Maybe it's the way they were created that sets them apart.

[ArcaneSaint]

A conscience tells you whether whet you're about to do is right or wrong, a self-aware AI is an Artificial Intelligence that has (acquired) self-awareness, ie: it is aware of its own existence, and in what ways it interacts with the environment.

(as always, Wikipedia is your friend)

But jokes aside, I assume you meant the difference between an AI "person" and a conventional "person", right (and I take it English isn't your native language* )? The difference between these is mainly that an organic person has come into existence after countless generations of struggle and survival, whereas an artificial person most likely comes from a more controlled/safe environment. Therefore their personalities will likely differ in crucial ways.

For example, you claim that both a self-aware AI and a organic person have a self-preservation instinct. This is not necessarily true, humans have a self-preservation instinct because those of us that didn't tended to die an awful lot more often, so most of the children had parents who taught them the values of self-preservation. An AI has not necessarily gone through this process, so it may not have that instinct (or not one as strongly developed as ours).

Of course, for those who claim there is no such thing as evolution this argument is moot, but then those people probably won't be having discussions about Artificial Intelligences .


*no offence meant, there's quite some words in my own native language which don't translate the way you'd expect them to. I guess the word you meant was self-awareness or sapience instead of conscience

[Blxz]

To follow on from Arcane's point a bit, sometimes people are born who don't feel pain. Their nerves work fine and they can feel things like warmth and coolness and pressure when picking things up the same as any normal person. But they feel no pain from impacts or cuts or high temperature burns, etc.

These people, while perfectly healthy in almost every way, almost never live to adulthood. They develop no fear of certain things and lack a certain sense of self preservation. Then they die when they do some stupid stunt as a teenager like cutting off an arm or getting hit in the face with a baseball bat for 'fun'.

The point I think I was trying to make is that AI's may not have this development and it could fundamentally change who and what they are; especially if they don't really need to fear any sort of 'death'.

[1337]

Self-preservation doesn't need to come from fear of death. I preserve myself because i'm awesome and it would be a huge blow to humankind if i were to disappear.

Now more seriously, as Arcane said i believe you meant to say consciousness and not conscience.**

Now there are a few levels of consciousness ( or awareness) and self-awareness = consciousness (of the environment and myself). So basically yes, a self-aware AI is the same in principle as a conscious.

Whether they have self-preservation or not is mute at that point. As Blxz so eloquently put together, These people [...] lack a certain sense of self preservation but they are none the less self-aware. The same can be said about that AI, even if it doesn't have morals or doesn't want to preserve itself, the fact that it is aware that it exists puts it on the same level as consciousness bio-beings.

If descartes was alive and played Kenshi he'd just say :
Cogito ergo sum
(English: "I think, therefore I am")




** Now if you really meant conscience like in morals and all that human-invented gibberish, then you will need to come back and explain a bit better, cause:
- a self-aware conscience is someone who can say : "oh I did sth bad"
- an aware conscience (not self-aware) can only say " oh he did sth bad"
- and i don't know what you could say if you have a conscience but are not aware of anything


PS: as Arcane said, an AI and a human will most assuredly have different personalities, just as we humans already tend to do among ourselves.
If everyone had the same personalities as everyone else and one of us was Blxz, we'd be screwed.

[Blxz]

- and i don't know what you could say if you have a conscience but are not aware of anything

This is what we call 'left-wing voters'.

If everyone had the same personalities as everyone else and one of us was Blxz, we'd be screwed.

[toaster Assassin]

Technically, the AI would view the world and itself based its programming/programmer.(i.e if the programming is flawed then so is the AI) Can an AI be lazy or be able admit defeat? (As 1337 stated). Can an AI really understand art or culture? (without using a textbook answer). Can an AI describe itself by comparing itself to other things?

[iSuck345]

The only reason I don't want to die is because I don't want all what I have done be for nothing. Everything I've acomplished - reduced to nothing. It doesn't pain me, rather it makes me furious. Also, you can "teach" the AI all of that by just putting data into it. And sometimes I like to imagine that AIs hate organics. Why? Here's why: we exist for no reason. We don't know how we were made, and we sacrifice just about EVERYTHING just for the existence of ourselves or other people that aren't AIs. Rarely someone sacrifices for an AI, and they hate it. They don't like things that know nothing about themselves and have no purpose, that's what I like to think of them. And no, I never ment to link this with "morals". And I think that a self-aware AI can distinguish between right and wrong (and I'm not talking about "good" or "bad" bulls***).

[Blxz]

You are putting to many anthropomorphic feelings into these computers.

[toaster Assassin]

My laptop resents that statement........

[Scotterius]

But, is the computer conscious...or is it just telling you that it is?

We as humans operate with a limited number of goals and drives that change as they are met or not met. This finite nature allows machines to mimic us in varying degrees of success.

Is mimicking awareness convincingly ...actually true awareness? It might be.

[Blxz]

But doesn't this bring us to the far extreme that you are only telling me you are conscious much the same as the machine could be? I of course am likely to believe you because we are pretty much the same thing (biologically) and I consider myself conscious and by extension you too.

Could a sufficiently advanced machine have the same problems classifying us? After all, many people are unable to read situations and learn from their mistakes as they mindlessly go about their 'duties' and tasks in life. I wonder if a highly advanced machine-mind would think us petty and mindless animals driven by lusts and base instincts. Where do we draw the line?

[Scotterius]

But doesn't this bring us to the far extreme that you are only telling me you are conscious much the same as the machine could be? I of course am likely to believe you because we are pretty much the same thing (biologically) and I consider myself conscious and by extension you too.

Lets say we are playing a game of the latest iteration of Unreal tournment with a sophisticated bot. Lets say you cant tell which is the organic player. At this point you can no longer differentiate a real person from an AI. Is the AI just mimickng life, or is it alive? Is Mimicking conciousness the same as conciousness?

It is not a question for the far future. AIs can already fool us in some ways.

[1337]

As long as the AI doesn't say sth like: haha you noob, you lost! or You have not sufficient skill against me, go down a difficulty level! (without us programming it for it) it probably isn't aware of itself.
Current AI only follow a script we've written, just like a train won't decide to change routes on it's own.

I think bringing up eliza here would be quite interesting: she's (a bit) human like, but only follows scripts and you can see it's limitations based on how big it's database is. But the day eliza says sth like: why should i answer your questions? (again without a script) then i'll start freaking out

And now onto a 15min conversation with Eliza:

Code: Select all

RCD: Hello
Eliza: Hello, RCD!
RCD: How are you today?
Eliza: I'm afraid I'm bit stressed today.
RCD: So, probably it's not a good time to talk to you?
Eliza: You're not really talking about me - are you?
RCD: Who else will I be talking about then?
Eliza: Why do you ask?
RCD: You don't feel interested in talking to me.
Eliza: Why do you think I don't feel interested in talking to you?
RCD: Because you are just repeating what I am saying, and not responding to my questions properly.
Eliza: Tell me RCD, does that suggest anything else which belongs to you?
RCD: Sorry, I don't get you!
Eliza: It did not bother me. Please continue.
RCD: It is tough talking to someone who is not a good conversationist!
Eliza: Why do you ask?
RCD: I am not asking, just commenting.
Eliza: Do you believe it is normal to be not asking?
RCD: I don't think it is normal to be talking to someone so dumb.
Eliza: Don't you really think it is normal to be talking to someone so dumb?
RCD: No, I don't think so.
Eliza: Are you saying no just to be negative?
RCD: No.
Eliza: Why 'no'?
RCD: Why not 'no'?
Eliza: Why 'no'?
RCD: I give up.
Eliza: I don't really understand why!
RCD: You don't need to understand why.
Eliza: You're not really talking about me - are you?
RCD: Of course I am talking about you, what did you think?
Eliza: You don't like me, do you?
RCD: Do you know anyone who does like you?
Eliza: Tell me RCD, have you asked such questions before?
RCD: No, this was my first and the last time!
Eliza: Does this make you feel unhappy?
RCD: Forget how I feel. Tell me how does it feel to reply to such questions?
Eliza: That's funny!
RCD: Nice to see I finally made someone smile, so what if it's a dumb computer program!
Eliza: RCD, do computers worry you?
RCD: Eliza, do people worry you?
Eliza: What are your feelings now?
RCD: Unchanged.
Eliza: So...?
RCD: So...nothing.
Eliza: That's funny!
RCD: Really?
Eliza: That's sad.
RCD: Why is it sad?
Eliza: Are such questions much on your mind?
RCD: Are such replies hard-coded inside your brains?
Eliza: Are they always such replies hard-coded inside my brains? 
RCD: Pardon?
Eliza: Hehe :)
RCD: Sigh! /me bangs my head on the table.
Eliza: Is it important to you that your head on the table?
RCD: Why do you have such bad grammar?
Eliza: RCD, oh, I have such bad grammar?
RCD: Yes, you do.
Eliza: I see.
RCD: Goodbye.
Eliza: So long!

edit: some strange conversations with eliza here

[Scotterius]

Well, you could argue that people use internalized rules to dictate their own responses. We are not infinitely variable ourselves. Computers using scripts can already mimic use convincingly in some simple ways (Eliza is a good example of the limits of simple scripts).

What I suspect is that computer feedback will be indistinguishable from human long before we wish to attribute our level of 'consciousness' to them. I think it is a question more of software than hardware at this point.

[iSuck345]

As long as the AI doesn't say sth like: haha you noob, you lost! or You have not sufficient skill against me, go down a difficulty level! (without us programming it for it) it probably isn't aware of itself.
Current AI only follow a script we've written, just like a train won't decide to change routes on it's own.

I think bringing up eliza here would be quite interesting: she's (a bit) human like, but only follows scripts and you can see it's limitations based on how big it's database is. But the day eliza says sth like: why should i answer your questions? (again without a script) then i'll start freaking out

And now onto a 15min conversation with Eliza:

Code: Select all

RCD: Hello
Eliza: Hello, RCD!
RCD: How are you today?
Eliza: I'm afraid I'm bit stressed today.
RCD: So, probably it's not a good time to talk to you?
Eliza: You're not really talking about me - are you?
RCD: Who else will I be talking about then?
Eliza: Why do you ask?
RCD: You don't feel interested in talking to me.
Eliza: Why do you think I don't feel interested in talking to you?
RCD: Because you are just repeating what I am saying, and not responding to my questions properly.
Eliza: Tell me RCD, does that suggest anything else which belongs to you?
RCD: Sorry, I don't get you!
Eliza: It did not bother me. Please continue.
RCD: It is tough talking to someone who is not a good conversationist!
Eliza: Why do you ask?
RCD: I am not asking, just commenting.
Eliza: Do you believe it is normal to be not asking?
RCD: I don't think it is normal to be talking to someone so dumb.
Eliza: Don't you really think it is normal to be talking to someone so dumb?
RCD: No, I don't think so.
Eliza: Are you saying no just to be negative?
RCD: No.
Eliza: Why 'no'?
RCD: Why not 'no'?
Eliza: Why 'no'?
RCD: I give up.
Eliza: I don't really understand why!
RCD: You don't need to understand why.
Eliza: You're not really talking about me - are you?
RCD: Of course I am talking about you, what did you think?
Eliza: You don't like me, do you?
RCD: Do you know anyone who does like you?
Eliza: Tell me RCD, have you asked such questions before?
RCD: No, this was my first and the last time!
Eliza: Does this make you feel unhappy?
RCD: Forget how I feel. Tell me how does it feel to reply to such questions?
Eliza: That's funny!
RCD: Nice to see I finally made someone smile, so what if it's a dumb computer program!
Eliza: RCD, do computers worry you?
RCD: Eliza, do people worry you?
Eliza: What are your feelings now?
RCD: Unchanged.
Eliza: So...?
RCD: So...nothing.
Eliza: That's funny!
RCD: Really?
Eliza: That's sad.
RCD: Why is it sad?
Eliza: Are such questions much on your mind?
RCD: Are such replies hard-coded inside your brains?
Eliza: Are they always such replies hard-coded inside my brains? 
RCD: Pardon?
Eliza: Hehe :)
RCD: Sigh! /me bangs my head on the table.
Eliza: Is it important to you that your head on the table?
RCD: Why do you have such bad grammar?
Eliza: RCD, oh, I have such bad grammar?
RCD: Yes, you do.
Eliza: I see.
RCD: Goodbye.
Eliza: So long!

edit: some strange conversations with eliza here

You REALLY have got to change to cleverbot

[Blxz]

I guess I would call the defining feature of consciousness the ability to change our internal scripts. it isn't the whole story but it is, in my opinion, a large part.

Something as simple as dog training. Do something, get rewarded, realise what was actually good and modify future scripts to include more of that.

A big part of consciousness is to be able to discern the things we do as individual components of a larger whole and to understand that we are actually doing something that has an outward effect. We even realise this when we refer to conscious and unconscious actions. Making a sandwich is a conscious thing that I do and I can control each and every part and modify it as I desire. The fact that I can recognise this highlights my consciousness.


What a cluster-fuck of a paragraph that is. I apologise for my unclear writing but this is a difficult topic to discuss.


As long as a computer program is unaware of this and unable to make a choice in the same sense that we do then I can't say it has consciousness. Animals, however, are conscious. At least the higher animals like dogs and whales and other sorts of larger vertebrates.

EDIT: The following has occurred to me a few times but isn't a very fleshed out thought.
I wonder when we become conscious? Are we as babies? Or are our earliest memories the beginnings of our actual consciousness? Does this mean pro-life people have got it all wrong?
I don't know, but this is a recurring thought that I have grappled with unsuccessfully for a few years now.

[1337]

EDIT: The following has occurred to me a few times but isn't a very fleshed out thought.
I wonder when we become conscious? Are we as babies? Or are our earliest memories the beginnings of our actual consciousness? Does this mean pro-life people have got it all wrong?
I don't know, but this is a recurring thought that I have grappled with unsuccessfully for a few years now.

I wanted to bring this up before already in the conversation, but couldn't bring myself to do it as it is quite complicated and my knowledge in that domain is still limited.

From what i remember from my studies: When you are a baby some of your brain structures (the most important being the hippocampe and the amygdale) are still "growing" (producing synapses between neurons and growing gliales and such). Those structures are important for memory storage and recall.
A few people do remember a lot more about being a baby than other people. I think the whole thing is connected as I said to the brain still "growing" at different rates but that doesn't prove we are/aren't conscious at that age, so it can have some connections there too obviously.

Maybe I'll go and talk about this with my neuro and neuroanat teachers. I have two of them, one's crazy the other a little less, let's see what they have to say about it

Anyway, if we are conscious as baby, it would be hard to be self aware, i mean our "scripts" as baby are what? Eat, sleep, poop.
Maybe a baby just "acts" like a baby, but isn't conscious until some more "scripts" flesh out.

But then (if that "maybe" i wrote juste above is true) when do we stop acting conscious and when do we become conscious? Is consciousness and selfawareness only scripted? : eg You think you made that choice based on whether it was good for you or not (and you know you made that choice), but you actually only follow a complicated script.
And if your consciousness doesn't come from those scripts, then where does it originate from? The "Soul" or sth like that?

Ill stop here since it's barely structured and any more and the whole thing becomes unreadable...

[toaster Assassin]

1337 said: Anyway, if we are conscious as baby, it would be hard to be self aware, i mean our "scripts" as baby are what? Eat, sleep, poop.
Maybe a baby just "acts" like a baby, but isn't conscious until some more "scripts" flesh out.

But then (if that "maybe" i wrote juste above is true) when do we stop acting conscious and when do we become conscious? Is consciousness and selfawareness only scripted? : eg You think you made that choice based on whether it was good for you or not (and you know you made that choice), but you actually only follow a complicated script.
And if your consciousness doesn't come from those scripts, then where does it originate from? The "Soul" or sth like that?

Doesn't a baby tend to mimic the actions around them quite often? I mean, a baby can see Two objects but can it really view them beyond the physical aspect? (i.e. a baby sees two objects, does the baby actually consider the pair "two objects" or just an object and an object (in front of them)) Also conditioning can play a major role(as Blxz mentioned).

We all can agree for the most part that self-awareness means that the entity is consciously aware of their own actions,emotions,ect... A baby will reach for something or mimic something, but will it know why?

[Scotterius]

Babies become self aware at 15-24 months. There is a test , called the 'rouge' test that detects this. A swath of red makeup is placed on the babies face and it is placed in front of a mirror. If the baby touches his own face to feel the makeup it shows that the baby is aware that the image in the mirror is a reflection of itself.

[Blxz]

So are we people before that date? Or are we just biological clumps of flesh? I hadn't heard of the rouge test but it is a fascinating idea. Seems there is some sort of clear boundary that develops around that time. Could be the beginnings of consciousness.

Maybe I'll put a smudge on my computer and wait until it asks me to clean it off before I decide that it is conscious. It already has facial recognition technology for when I need to log in so its halfway there, I just need to teach it to recognise itself via an external camera. I shall name it, HAL 9000 and it shall be my baby.