Ethics & destiny of AI creations

[shunyadragon]

Sounds like satire though.

With a heavy dose of sarcasm.

[JohnnyP]

If they can 3D print a pizza why not a mind?

NASA Funds 3D Pizza Printer

More to the point:

The first 3D printed organ -- a liver -- is expected in 2014

[shunyadragon]

If they can 3D print a pizza why not a mind?

NASA Funds 3D Pizza Printer

More to the point:

The first 3D printed organ -- a liver -- is expected in 2014

Interesting . . . no comment at present.

Will they be able to do the whole body next? Beam me up Scotty!

[David Hayward]

Here's an 'African' theological perspective.

[Sparko]

The equivalent intelligent being most likely will be neither the mirror of a human brain nor anything like a computer as we know it.

If it is not like our mind then how can you judge it an artificial intelligence? I would have to mimic our mind in order for us to recognize it as a real mind and not just some expert database system.

One hypothetical way it would develop is an equivalent to the neuron and neural networks that may be integrated into an equivalent to systems of neuron networks equivalent to a brain.

That doesn't even make any sense. If you are saying it would mimic our neural net, then you are back to my initial objections. We don't know much about how an actual mind works. We know some pieces here and there but there just isn't any way we can observe an entire mind at work, much less try to duplicate it in software. Just having a large neural net doesn't equal "mind" or intelligence. An elephant has a larger brain and neural network than we do and it doesn't have the intelligence of a human being. It's not just a matter of size of neural connections, but of how they work together. And we just don't know how that works. The neurons don't just randomly connect and fire and you have a brain.

[Sparko]

If they can 3D print a pizza why not a mind?

NASA Funds 3D Pizza Printer

More to the point:

The first 3D printed organ -- a liver -- is expected in 2014

A liver is a lot simpler than a mind. And they are starting with liver cells and just laying them down.

[shunyadragon]

A liver is a lot simpler than a mind. And they are starting with liver cells and just laying them down.

We may have a parallel if it is neurons they are starting with then series of neuron networks,' and just laying them down.' The problem is they probably would not have the memory patterns imprinted on the system of neuron net works. The would likely end up with a 'tabla rasa.'

[Sparko]

We may have a parallel if neurons they are starting with, 'and just laying them down.'

Our brain has some pretty complex structures and different types of neurons. Just laying down a sheet of them won't cut it.

I don't know if we can create a truly self-aware artificial mind or not - some day. But if we do, I would think it would deserve the status and rights of a real person. At least until it turns into Skynet and tries to wipe us all out with terminators and hunter-killer drones.

[shunyadragon]

Our brain has some pretty complex structures and different types of neurons. Just laying down a sheet of them won't cut it.

I don't know if we can create a truly self-aware artificial mind or not - some day. But if we do, I would think it would deserve the status and rights of a real person. At least until it turns into Skynet and tries to wipe us all out with terminators and hunter-killer drones.

Different types of neurons can be dealt with in this methodology, but again I consider this extremely hypothetical by the present technology.

[David Hayward]

This paper, "Fractals and the irreducibility of consciousness in plants and animals" -- click on "PDF preview" to see the summary -- claims that "In animal nervous systems electrical signalling is a fractal process..." therefore, because of the 2nd law of thermodynamics, "... it is impossible to extrapolate backwards to determine the rule or rules that govern consciousness. Thus any attempt to create a model of consciousness is doomed to failure."

Which I take to mean that the paper's author is claiming that it is not possible in theory to discover exactly how the human mind works, and to model (or replicate) it, because the relevant rules/algorithms by which it works are in principle unknowable.

[shunyadragon]

This paper, "Fractals and the irreducibility of consciousness in plants and animals" -- click on "PDF preview" to see the summary -- claims that "In animal nervous systems electrical signalling is a fractal process..." therefore, because of the 2nd law of thermodynamics, "... it is impossible to extrapolate backwards to determine the rule or rules that govern consciousness. Thus any attempt to create a model of consciousness is doomed to failure."

Which I take to mean that the paper's author is claiming that it is not possible in theory to discover exactly how the human mind works, and to model (or replicate) it, because the relevant rules/algorithms by which it works are in principle unknowable.

I am very suspicious in science when some one concludes 'impossible,' unknowable' or 'doomed to failure.' Yes and no, animal nervous systems electrical signaling appear in a fractal pattern, but I believe it is a mistake to call it 'a fractal process.' Virtually everything, all series of events, in nature follow a fractal pattern, but again they are not a 'fractal process.' Chaos math is not a problematic enigma, it is simply a descriptive nonlinear math pattern as a series of events occur when there are many variables.

The article lacks substance behind the claim, based on what I read. If science developed a working neuron network, it would function in a fractal pattern.

[David Hayward]

Yes and no, animal nervous systems electrical signaling appear in a fractal pattern, but I believe it is a mistake to call it 'a fractal process.' Virtually everything, all series of events, in nature follow a fractal pattern, but again they are not a 'fractal process.' Chaos math is not a problematic enigma, it is simply a descriptive nonlinear math pattern as a series of events occur when there are many variables. ...If science developed a working neuron network, it would function in a fractal pattern.

I think the point of the paper I linked to is that if a working artificial neuron network functioning in a fractal pattern were to be developed, it wouldn't be working in the same pattern as our brains do, because that pattern is (it claims) unknowable.

I am very suspicious in science when some one concludes 'impossible,' unknowable' or 'doomed to failure.

I too am suspicious of 'unknowable', etc; a healthy scepticism is, well, healthy. Also, although the paper claims/premises that neuron networks function in a fractal pattern, that is new information to me (though that might simply reflect my ignorance) -- should the paper's premises be invalid, that will invalidate the paper's conclusions.

[shunyadragon]

I think the point of the paper I linked to is that if a working artificial neuron network functioning in a fractal pattern were to be developed, it wouldn't be working in the same pattern as our brains do, because that pattern is (it claims) unknowable.

The pattern would be fractal, not unknowable. The design of the neurons and neuron neural network would determine the function and not the fractal pattern of the events.

I too am suspicious of 'unknowable', etc; a healthy scepticism is, well, healthy. Also, although the paper claims/premises that neuron networks function in a fractal pattern, that is new information to me (though that might simply reflect my ignorance) -- should the paper's premises be invalid, that will invalidate the paper's conclusions.

Not news to me. The premises of the paper are faulty. Believe it or not all the events in nature follow a fractal pattern. Every math formula that has many variables has a nonlinear fractal pattern in the results, Just like 'no two snowflakes are alike, all snowflakes will look like and act like snowflakes,' 'No two brains will function exactly alike, but all brains will function like brains. It will be the design of the artificial brain that will determine how the brain functions, and not the fractal pattern of the neuron network activity.

[Sparko]

another question...

Would a true artificial mind/brain be a psychopath? It would have no emotions (no hormones or endocrine system) and would be pretty much unable to empathize with others because empathy requires emotion, and being able to place yourself in someone else's position.

Creating a completely digital personality could lead to creating a uber powerful psychopath that has control over our entire society and technology. A real skynet.

kinda scary

[shunyadragon]

another question...

Would a true artificial mind/brain be a psychopath? It would have no emotions (no hormones or endocrine system) and would be pretty much unable to empathize with others because empathy requires emotion, and being able to place yourself in someone else's position.

Creating a completely digital personality could lead to creating a uber powerful psychopath that has control over our entire society and technology. A real skynet.

kinda scary

You're jumping the gun here, first the artificial brain would not be a digital brain or personality. It would be based on an artificial neuron network equivalent to a natural brain. Your making a lot of assumptions of unknowns.