The difference between a self-aware AI and a conscience

[iSuck345]

Ok back to tue question: the difference between a computer who THINKS exacly like an adult human and a human who thinks like an adult human. That's what I was trying to ask. Now that I made it more clear, I shall enjoy seeing the others have their brains popping out and having scratch marks.

[Blxz]

The key word in your question is 'Exactly'. If a computer thinks EXACTLY like a human adult then it is conscious in the same way as that human adult. Only the hardware is different.

Of course getting that 'exact' bit to be true is the challenge and there is no way that a computer would think the same having 'grown up' in a different situation and with different influences. But lets pretend you just copied my brain into a computer. From every one else's point of view (except, most interestingly, mine) that computer is me. If I were to die everyone else would be happy dealing with that computer and I wouldn't be missing. The most unfortunate part of course is that I would be dead in that scenario. Hence copying over my mind isn't immortality for me personally.

The only way to do it properly would be a gradual transition where I simply begin to augment my brain partially with mechanical components while remaining conscious and then deactivating the relevant biological part. That way whatever makes up 'me' is still continuous and not simply a copy that no one else can distinguish.

[turbochop]

Now see, I both agree and disagree with this point. If I were given the option, I would decline to have my brain copied onto a computer. It may have my thoughts, likes, dislikes and personality, but it's NOT ME! It's my mind uploaded to a computer!

Once I die, my soul travels on, I don't just magically start living inside this machine. You could reference just about ANY media that involves cloning and you see the conflict of the clone living out the originals life, with the original going "WTF?! that's not me!". My point is, a mind uploaded to a computer is just data. It's not the original's conscious, it's artificial conscious, and the idea of death becoming trivial with the advent of cloning and mind uploads is perverse to say the least. Maybe I missed the mark completely, I dunno.

[iSuck345]

Up to now, I've seen turbo the only one to give his answer, instead of talking about the brain, how it develops and etc.

[toaster Assassin]

There is an older topic (that I posted a long time ago) where scientists placed some rat brain cells (or something similar) onto a microchip and then placed the microchip into a small robot. The rat cell microchip (lets just call it Wilfred for now) was able to interface with the robots equipment and not crash into walls.

Anyway, if the computer was able to exactly translate your essence into data (thus becoming "you) it can only predict (based on your pirated essence) how to act. The robot would act like ''traditional you'' under every circumstance, while you have the option to "change yourself" (favorite foods,ect..) and still remain you.


Something to think about:

If I remember there is an old saying: We are not born us, instead we become us.
Perhaps, we are just a "work in progress" (building up until we die), which would mean that the computer would just mimic you of that moment (And could not really mimic the dynamic you)
Or I really need sleep, I don't know.

[Blxz]

The cells that make up you are dying and being remade constantly to the point that not a single cell alive in your body now was present at birth. But 'you' are still there. Basically the new cells take on whatever it is that you are, ie. they don't replace you and then they take on the same role.

This lends credence to my theory that it would be possible to gradually replace yourself (or your brain at least) with mechanical and electronic parts but still retain your mind and body. Basically as I said in my previous post, there would be no copying of yourself just a changing of the physical makeup.

Does that explain my thinking more accurately for you Turbochop?

[turbochop]

Yeah, I understand. That raises another question though: How much organic tissue can you replace and still retain your soul? Because without a soul you're just a shell. Would that mean that clones and minds uploaded to computers (I'll from now on call them Artificial Minds, or AM's for short) have no soul? What exactly makes up a soul? Isn't the mind \ conscious linked to the soul? If you replaced your brain with an AM, would your soul be relinquished? so many questions I think thats a good answer to the topic though: The difference between self-aware AI and a conscience is the presence of a soul.

[1337]

Guys you made a big error/assumption there!!

Neuron cells don't replace themselves after you are born!

You are born with a fixed amount and from then on, the number just goes down and you have to make the best of what you have.

This is true for the entire cortex. (except maybe the hippocampus, but it's very limited there too)

Just wanted to say this before you go down the wrong road to much

[iSuck345]

Thank god this isn't a school-based forum...

[demidevil]

Guys you made a big error/assumption there!!

Neuron cells don't replace themselves after you are born!

You are born with a fixed amount and from then on, the number just goes down and you have to make the best of what you have.

This is true for the entire cortex. (except maybe the hippocampus, but it's very limited there too)

Just wanted to say this before you go down the wrong road to much

yeah you can't regrow them, but there's studies that show you can transplant them, though essentially you're correct

to the question though the difference between a self-aware ai, and a conscience is that a human's conscience comes from chemical reactions within our body, whereas an ai's conscience would have to come from highly complicated formulas, and just like how an ai can never generate a truly random number, it would never generate a truly self aware conscience.

[iSuck345]

Guys you made a big error/assumption there!!

Neuron cells don't replace themselves after you are born!

You are born with a fixed amount and from then on, the number just goes down and you have to make the best of what you have.

This is true for the entire cortex. (except maybe the hippocampus, but it's very limited there too)

Just wanted to say this before you go down the wrong road to much

yeah you can't regrow them, but there's studies that show you can transplant them, though essentially you're correct

to the question though the difference between a self-aware ai, and a conscience is that a human's conscience comes from chemical reactions within our body, whereas an ai's conscience would have to come from highly complicated formulas, and just like how an ai can never generate a truly random number, it would never generate a truly self aware conscience.

So, ALL of the Sci-Fis who use robots that have a self-aware conscioussness AND act, think and learn like a human are wrong. Is that what you're saying?

[Mattk50]

to the question though the difference between a self-aware ai, and a conscience is that a human's conscience comes from chemical reactions within our body, whereas an ai's conscience would have to come from highly complicated formulas, and just like how an ai can never generate a truly random number, it would never generate a truly self aware conscience.

This is nonsense. The reactions in our brain are not fundamentally different from logic gates in a computer, there's no intrinsic difference between algorithms running on a biological or electronic computer. Because of this, higher level functions that emerge (such as a conscience) are not precluded by the different materials used in the logic. What you just said is equivalent to saying that because calculators run off electronic circuits, the human brain can't do math due to the fact it uses thoughts.

In addition, the determinist viewpoint says that nothing, not a computer nor anything else at all can generate a truly random number. The difference is that because we build computers, we understand all the details of how they work and so we can easily predict what number they will come up with. Many more "random" number generators use static noise in a room as an input for a computer to generate numbers, this makes it harder to predict the outcome however if you have all information about sound waves that will hit the mic in that circumstance, it becomes just as easy to predict. The same is true for anything generating supposedly random numbers.

[kiseruyoru]

The entire concept of causality (if applied to human action) would strongly suggest that a fully realized AI is entirely possible, and considering our technological progress - eventually will be a reality.

On an almost unrelated stance I would think that an AI is self-aware the moment it questions its own existence. I realize adults don't really do this, also pretty sure this is because the question is maddening, but I've never known someone who didn't do it as a child.

On a technical stance I'd say the 'true AI' would be a single program that allowed the acquisition of knowledge - this would (presumably) eventually lead to the question of one's own existence.

As for the random number thing being discussed here, the causality bit is my response to that. The 'random numbers' of a computer are FAR more random than those of a person. Computers can simply make far more calculations in the time it takes to pick a random number. Granted, we're never aware of most of the calculations we're making (at least according to many modern theories) but still not as many as computers can do, even if we take it on a purely mathematical level (numbers only, not converting words, concepts, etc to numbers) -- and even human err can be quantified and explained in a formulaic manner -- the most obvious proof of concept for that being Spoonerisms.

I wish I knew more about programming though - I strongly considered going to school for it so I could work in the field of AI -- but meh, too big a chance that I'd be meaningless to the pursuit or worse, it not even happen in my lifetime. I'm just not smart enough.

[Blxz]

Richard Dawkins quotes someone in one of his books (The Selfish gene) when he refers to consciousness. Basically he was talking about simulations that we run in our minds about the real world. We do it all the time whenever we think about any action at all or when we think about something to do. The only time when you are not simulating is basically when you are driving or doing somehting and your mind goes on autopilot and suddenly you think "how did I end up here".

Back to the point, Dawkins and the man he quoted were of the opinion that consciousness arises when our models of the world become so advanced that they include a simulation of ourselves within that model. Basically we can imagine ourselves doing something as opposed to an event happening. It is at that point, he argues, that we become both self aware and develop our term of consciousness as well.

Personally, I find it to be a possibility. (EDIT: And the post above mine seems to agree in principal. Wasn't there when I started to begin my wall of text though.)

Also, turbochop, putting religion to the side there is no such thing as a soul. To the religiously minded they attribute the soul to our immortal consciousness that can/will continue on in the event of our death. Scientifically speaking there is no basis for a soul nor any evidence and so shouldn't have any effect on merging our minds gradually into a computer; in fact if our soul is mutable and able to leave our bodies as the religious believe then it should simply transfer along with the rest of our consciousness at the same time. I will leave my further comments about whether I believe or not out of the discussion.


Finally, my reply to Leet presented in sexy link form rather than more walls of text:

http://biologybrainfood.wordpress.com/2 ... egenerate/

http://biology.about.com/od/Brain/p/Reg ... -Cells.htm

This one provides something different that shows growth of white matter as opposed to the actual circuitboard of the neurons (which is what I am suggesting we replace, the hardware). Plus, we all love yahoo answers:
http://answers.yahoo.com/question/index ... 006AAVvaHD

And finally a whole bunch of cool stuff about science that everyone should read because I said so. Only the 3rd one down is actually relevant to the discussion.
http://dsc.discovery.com/tv-shows/curio ... ctures.htm

[1337]

The yahoo answer isn't wrong but it doesn't relate to what we were talking about, since he's talking about myelinisation. There's a difference between a brain growing and regenerating

The other studies are not really that good actually and hasn't evolved that much since the start. Comparing a primate brain to a human brain wasn't a good idea (let's not even talk about a rat brain ). We have billions and billions neurons, about ten times more than a primate. We don't generate more neurons, because those billions of neurons form a complex circuit and adding to it will only degenerate that balance. We do get some in the hypocampus but that is only during our teen years and doesn't compare to the number that we lose during the same time.

The reactions in our brain are not fundamentally different from logic gates in a computer, there's no intrinsic difference between algorithms running on a biological or electronic computer

+1
Synapses use chemicals, but they basically only convey electronic impulses, the same as logic gates.

The only real problem i see (imo) is the complexity of the whole brain: over 80 billions of cells that are interconnected (neurones, interneurones, inhibitory neurones, ...), that is hard to reproduce to 100%. How much hardware would be needed to make that complex of an AI? Also how many code lines would we need to write?

[Blxz]

My point though is that the neurons are the hardware. The brain is able to grow matter in relation to memory and frontal cortex regions. If you could sync a chip of some sort into your brain to take on the role of your heart beat you could then remove that related part of the brain at a later date. Then you could do the same with a memory chunk when it is copied over. Then a little more. Do this over a few years and provided nothing is totally removed and the circuits sync up properly then there is no reason that we couldn't change the hardware of our brains to a physical piece of computer technology.