h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

I'm reading Roger Penrose's 'The Emperors New Mind' and its quite complicated but interesting, and its giving me loads of ideas for guide entries, so watch this space.
And id recommend the book to anybody, especially if you stongly agree/disagree with AI viewpoints.

Hello SuperSam,

Found myself here after reading you short story and this journal entry caught my attention.

Did you mean all viewpoints, or AI (artificial intelligence) viewpoints?

I ask for two reasons; 1) if you mean AI, this interests me as I used to be an AI student, please tell me more about the book! 2) if you mean all viewpoints, this also interests me...

Can someone disagree with all points of view? I would argue not; lets take the following situation... Person D thinks something is false, person E thinks that same thing is true. Person F disagrees with all points of view, therefore he thinks that this thing is (D) not false, so true and (E) not true, so false...

A paradox occurs and the world ends?

Sorry, any excuse for an argument, but i look forward to your reply.

Yours,

Discus.

I meant AI viewpoints, and the book is about trying to prove that computers can never replicate computers, no matter how advanced because the experience of being alive, and feeling and thinking can never be in a computer(or robot for that matter), because these things can't be expressed by an algorithm, well thats where I am so far, penrose uses stuff like turing machines and churchs lambda calculus to explain and it gets quite confusing(for me anyway), and where i am now hes using complex numbers.

As for disagreeing with all viewpoints, or agreeing which would have the same affect, you are assuming that person F couldn't agree with a viewpoint that didn't yet exist, so couldn't express his viewpoint, merely disagree with everyone else.(if that makes any sense at all)


Sam

Ah-ha... but when this elusive viewpoint is conceived, it will be a viewpoint, so person F will disagree with it... I propose that it is impossible to disagree with all viewpoints, as per my argument above.

Regarding AI machines emulating human behaviour, one must ask what it is that creates this behaviour in humans. For example, if you ask a priest what causes our intelligence, you will get a (probably) very different answer than if you asked a psychoanalyst.

There is a very powerful scientific argument that indicates that human intelligence is encapsulated by the brain; that there is no spirit, so to speak. The brain is merely a network of neurons relaying electronic signals. You could, possibly, create a neural network that would emulate this, presumably resulting in a rather clever machine. Does your book cover neural networks?

Yours,

Discus

youre right u cant disagree with all viewpoints(no point flogging a dead horse, as i never say)

you're comparing neurons to computer programs or components, which i think is a dangerous analogy (please, protect me from those murderous analogies), if you took apart a computer, you would get a pile of useless parts (on their own) and there is nothing in there that would b able to think, if you took your brain apart into neurons, granted those neurons would be pretty useless, but you could actually keep those neurons alive, in the right conditions, for a while anyway, and the neurons would have the capacity to take in minerals and ions so it could remain alive for as long as possible, that neuron would be thinking for itself if you like, and this separates it from the computer components, now taking the computer an brain whole, the computer can only make decisions based on the programs it has, it has no capacity for original thought unlike the brain even if it has the same structure as a brain.

And if the computer could be intelligent, then you would have to consider it alive and the computer, if doing its task (as specifed by Turing) to replicate a human, the computer would have feelings, could you ask a computer to perform a task if that meant you were basically treating it like a slave, thats where ethics and morality comes in, the computer, essentially would be human in all but name, it has become more than simply a machine used as a tool, it would be able to survive on its own, and when running out of energy would be able to go to the nearest power socket and charge up, hunter gatherer style, now is the pursuit of AI worth it, comsidering it may well be impossible and even if not, ethically questionable?

SuperSam

Actually, I was not comparing neurons to computer components, but an actual piece of software.

A neural network simulates (almost perfectly, actually) the way that neurons connect to each other and communicate in the brain. If you wre to remove some nodes from this neural network, it would be less powerful, but it would still function.

If you have a machine powerful enough, you could use a neural network to precisely replicate a human brain - and supply it with similar inputs that we get from our nervous system.

The one flaw here is that the way neurons operate can (i believe) be affected by the contents of the blood stream (such as alchohol, testosterone etc.) and a neural network cannot....

Yours,

Discus.

also, different levels of chemicals can actually make the brain work in a completely different way, effectively 'rewiring' it, ill take ur word for the rest of it, sry 4 misunderstanding ur analogy

SuperSam

No probs. Interesting discussion, I'll keep an eye on your entries in future.

Looking forward to more.

Yours,

Discus.

ps. Thanks for your comments re: The Excuse.