technological singularity

[some dumb kid]

i was at the library yesterday and i saw a book on the shelf called "rapture for the geeks: when AI outsmarts IQ" by richard dooling. thinking it would be an interesting read about the far flung future i got went home and read it to my horror i found out that this "technological singularity" when computers become smarter than humans was going to ocur very soon as early as 2015! the idea of the earth run by a giant super computer really scares me! also the book talks about "transhumanism" wich is basicly trasnporting your mind in to a computer that idea also spooks me i like my body and i dont want to live for ever. any one else know anything about this event that they could tell me?

[PraedSt]

You get spooked waaay too much SDK.

I think this topic has been covered before on here (you'll probably find it with a quick search), and the conclusion reached was 'not very likely'.

[some dumb kid]

yea thats me i have a nasty tendency to find things that scare me without even trying, my serch turned up about 7 topics wich one would you recomend?

[PraedSt]

I'll have a look. But in the meantime maybe start off with the newest one and work backwards?

[PraedSt]

Here you go.
http://www.bautforum.com/questions-a...ngularity.html
Fifth post down (01101001) also gives the list of all other threads.

[some dumb kid]

some interesting arguments on both sides but nothing really convincing anyone have any opinions that arnt in the previous topics

[PraedSt]

You're back.

Yes well- we're dealing with a) the future, and b) random little things called humans, so you're never going to know really. Not until the machines are running the show. Cheer up! Might not be so bad.

I'm sure Swift will be soon be along to say: 'I, for one, welcome our new Machine Overlords!'

[mike alexander]

I, for one, welcome the Singularity.

[ravens_cry]

I. for one. think the idea of Singularity is so much horse apples. Not the idea of intelligent computers, or even the idea of a computer that is more intelligent then a human. After all, we can build a machine that can lift more then a man, so why not a machine that can think more?
But this Singularity has been promised, and pushed back and prophesied and pushed back, so much it is just silly. Sure, in fiction AI seems to be a major crapshot, full of handwaving and Phlebotinum. Positronics, anyone? But that is because the authors have no idea how to produce an strong AI. If they did, they would be writing the code, not writing the book.
But if we actually knew how make a mind in a computer, we would be able to control all the details. We could make it completely loyal, and enamored with its function. Sure, there can and will be bugs, but it won't make it 'turn evil'. Evil is a rather difficult program in itself, it isn't going to be caused by some slightly mangled computer code. It certainly won't make the eyes turn red.

[PraedSt]

I, for one, welcome the Singularity.

Thanks Mike.

Saved me from our new punctuation Mod.

But if we actually knew how make a mind in a computer, we would be able to control all the details. We could make it completely loyal, and enamored with its function. Sure, there can and will be bugs, but it won't make it 'turn evil'. Evil is a rather difficult program in itself, it isn't going to be caused by some slightly mangled computer code. It certainly won't make the eyes turn red.

Isn't the fear based on the fact that, sooner or later we'll make an AI that's self-aware? A peeved self-aware AI could go and make an evil AI, for example.

[captain swoop]

I am self aware, i fix digital Xray equipment, how would I make an evil AI if I was peeved?
Why is there a supposition that a self aware AI will be able to program another one? Wouldn't it depend what the AI was built to do?

[ravens_cry]

Thanks Mike.

Saved me from our new punctuation Mod.


Isn't the fear based on the fact that, sooner or later we'll make an AI that's self-aware? A peeved self-aware AI could go and make an evil AI, for example.

Not if we make it so it won't. Why program being peeved in anyway? Unless it serves a use for the final mind, Why add it? For that matter why even give it programming skills? Just because it it self is an AI, would no more make it an expert on artificial intelligence, then being a meatbag wetware makes Joe the plumber an expert biologist and neurologist.

[PraedSt]

Not if we make it so it won't. Why program being peeved in anyway? Unless it serves a use for the final mind, Why add it? For that matter why even give it programming skills? Just because it it self is an AI, would no more make it an expert on artificial intelligence, then being a meatbag wetware makes Joe the plumber an expert biologist and neurologist.

True, true. Just a few hypothetical questions though...
Isn't part of the technological singularity based upon an exponential increase in AI function? A good way of achieving that would be to build minds, who can build minds, who can build minds, who can...
Also, what's to stop them collaborating? An expert AI biologist + an expert AI robotocist + an expert AI programmer, etc, etc.

I like the idea of hyper-intelligent Minds, I have to say- like in the Culture novels. They have a sort of symbiotic relationship with humans if I remember correctly. No guarantee of a 'no-evil' Mind, but they're so intelligent that they've decided not to bother.

[stutefish]

Who's to say a capacity for peevishness isn't a necessary condition for some other desireable features?

Speaking of desire, what is peevishness a sense of thwarted desire?

We could, of course, seek to perfect AI zen masters, which desired nothing, but then why make them AIs at all, instead of just big dumb computers like we already have?

A machine that desires, on the other hand, is a machine with an urge to solve problems, to go beyond its basic programming, to innovate and adapt and persevere. Such a machine, with sufficiently versatile hardware, would be able to accomplish far more work, with far less direct supervision, than any robot yet built. It would be ideal for exploring Mars or the Marianas Trench, or extracting mineral resources from Siberia, or doing any number of other useful but difficult things where sheer human refusal to give up would be an asset.

That's where I see the big AI breakthrough: Not machines that think like humans, but machines that desire like humans.