How close is AI?

[samk]

"No, I don't have a clue why they're doing that stuff and I can't tell them not to, but it works 99.99% of the time" is also a good description of humans only a scratch beneath the surface of consciousness

And as far as the internet waking up goes, it is best thought of not as a sleeping giant, but a giant petri dish waiting for the right spore...
a mindspore drifting out from a rogue basement laboratory on an open cable

 

[8ball]

It depends on what you mean by "left behind". We're already in a situation where the complex inter-dependencies of, for example, managed network devices in telecommunications companies can't really be understood by people . . .

What, ANY people??

Anyway, what I meant wasn't that most or all people don't understand the workings of the systems any more, but that we'll be intellectually left behind, superceded and discarded by these systems. I think we'll use their strengths and they'll use ours, to the point where the overlap becomes such that distinguishing between the two kinds of system becomes academic.

 

[merlin wood]

At the moment I think AI will be practical, once we can make a synthetic copy of the human mind.

But then you'd have to find out where and how there's a human mind that you could copy.

 

[fudgefactorfive]

I was reading about Descartes recently - he lived at a time when there was a trend for creating clockwork automata for entertainment purposes - he said he couldn't see any difference between a clockwork limb, with its pistons and ratchets and wheels, and a human limb with its muscles and fibres and layers. People were making claims about thinking machines as far back as Bacon. I get the same vibe from people who think we're on the brink of an AI age.

Even if through sudden breakthroughs, we can simulate the structure of a mind, I think there'll be a few hundred years of creating terminally insane entities that shriek in pain for fractions of a second before dying, and dying again, and dying again, and dying again, first. Presuming we can maintain the industrial base to support such research, which is quite unlikely imo ...

 

[8ball]

I think there'll be a few hundred years of creating terminally insane entities that shriek in pain for fractions of a second before dying, and dying again, and dying again, and dying again, first....

Then someone will get the bright idea of not playing Westlife in the lab and we'll be ok

 

[samk]

Even if through sudden breakthroughs, we can simulate the structure of a mind, I think there'll be a few hundred years of creating terminally insane entities that shriek in pain for fractions of a second before dying, and dying again, and dying again, and dying again, first.

No, they'll just be like max_freakout

 

[laptop]

I was reading about Descartes recently

Book recommendation?

 

[fudgefactorfive]

Book recommendation?

The same Fritjof Capra one I was going on about recently ... "The Web of Life" - layman's guide to cognitive science, cybernetics, neural networks, physical networks, chaos theory and attractors in phase space, ecosystems etc. IIRC there was a section about AI in it, about a decade out of date though.

lol and lol at 8ball and samk's threads btw

 

[laptop]

Fritjof Capra

Ah

 

[rich!]

TeeJay, you're not getting it. This ain't science baby, this is engineering. As for spontaneously generating, no. This won't happen in someone's garage. Even little expert systems can chew through CPU time at a rate you wouldn't belive.

Well, it might happen in someone's garage...

But, at the moment, the state of the art is that it's pretty damn near impossible to work out how to pick something up from scratch. Being able to do anything with it...

 

[merlin wood]

I was reading about Descartes recently - he lived at a time when there was a trend for creating clockwork automata for entertainment purposes - he said he couldn't see any difference between a clockwork limb, with its pistons and ratchets and wheels, and a human limb with its muscles and fibres and layers. People were making claims about thinking machines as far back as Bacon. I get the same vibe from people who think we're on the brink of an AI age.

Even if through sudden breakthroughs, we can simulate the structure of a mind,....

But then suppose the mind is Cartesion and to the extent that it is an immaterial subject of experience that is dostinct from the body and is required for consciousness to be at possible?

 

[fudgefactorfive]

But then suppose the mind is Cartesion and to the extent that it is an immaterial subject of experience that is dostinct from the body and is required for consciousness to be at possible?

I don't think Descartes was qualified to lecture you or I on neural networks:

We see clocks, artifical fountains, mills and other similar machines which, though merely man-made, have nonetheless the power to move by themselves in several different ways ... I do not recognise any difference between the machines made by craftsmen and the various bodies that nature alone composes.

 

[Jonti]

The point Descartes is making there is that the body can be understood as a machine. The way the body works can be understood in exclusively material terms.

The advances in medical science made since his time are based on this idea -- it has been enormously beneficial and fruitful in advancing our understanding of how the body works; how to fix things when it goes wrong; and how to cure and prevent disease.

Descartes was spot on, to the extent that we can, and do, understand the body as a machine.

 

[fudgefactorfive]

The point Descartes is making there is that the body can be understood as a machine. The way the body works can be understood in exclusively material terms. ...

Descartes was spot on, to the extent that we can, and do, understand the body as a machine.

no, that's nonsense - the body is not a linear, clockwork, mechanical machine

 

[ATOMIC SUPLEX]

How close is AI?

Ai is in the waterloo init, where are you and we can work it out.

 

[Jonti]

At the moment I think AI will be practical, once we can make a synthetic copy of the human mind. Beyond that, may be something radically different will appear, in the same way birds are different from planes, planes might not be as efficent but they're a lot faster.

Semiconductor technology, Intel's lastest chip has 582 million transistors. Wikipedia suggests there's 100, 000 milion Neurons in the human brain.
Number of transistors double every 18 months.
So....................................

You're making a heck of an assumption by positing ...

[CENTER]neuron <=> transistor[/CENTER]

What are the grounds for suggesting such an equivalence?

 

[likesfish]

if we did build an AI it would be awfully lonely around here
well you seen any signs of intelligence around here?

 

[Bob_the_lost]

You're making a heck of an assumption by positing ...

[CENTER]neuron <=> transistor[/CENTER]

What are the grounds for suggesting such an equivalence?

Do you think there is something special about neurons or that neurons are merely more complicated than transistors (which they are).

 

[Jonti]

More complicated is putting it mildly, isn't it?

It's plain that neural cells have an enormously complicated internal structure; and that there is a range of phenomena at different scales, mediated by different forces and influences. Sure, someone can claim all that stuff is irrelevant to the question of what AI is, and that it all boils down to be equivalent to a light switch (well, OK, transistor). But in the absense of any evidence for such an equivalance there is no particular reason to entertain the notion.

It seems unlikely that the neuron is the simplest way nature has of making a binary switch (transistor). Why, then, all that other stuff?

 

[Bob_the_lost]

More complicated is putting it mildly, isn't it?

It's plain that neural cells have an enormously complicated internal structure; and that there is a range of phenomena at different scales, mediated by different forces and influences. Sure, someone can claim all that stuff is irrelevant to the question of what AI is, and that it all boils down to be equivalent to a light switch (well, OK, transistor). But in the absense of any evidence for such an equivalance there is no particular reason to entertain the notion.

It seems unlikely that the neuron is the simplest way nature has of making a binary switch (transistor). Why, then, all that other stuff?

It's not a simple 1:1 connection, which meant it'd take more than 15 years or so to hit the 100,000 Million transistor mark to get equality of function. Still...


Still, let's say it's only a thousand times more complicated than a transistor, or that a thousand transistors is needed to replicate a single neuron's function. Just stick a thousand chips together.

 

[Jonti]

I'm with gurrier, that we already have AI. It is very specialised, and it works in the dark, so to speak, without the awareness or creativity that some posters here have explicitly associated with intelligence. But it is all over the place in the modern world. Depending, of course, on what we mean by "intelligence".

Some vocabulary control will likely help the debate along. One widely accepted definition of intelligence is "the faculty to acquire and apply knowledge".

 

[Jonti]

... let's say it's only a thousand times more complicated than a transistor, or that a thousand transistors is needed to replicate a single neuron's function. Just stick a thousand chips together.

Fair enough, as far as it goes, provided all the phenomena involved in the functioning of a neuron can be modelled with binary switches.

As I've said above, the faculty to acquire and apply knowledge is something that machines already have.

 

[laptop]

Depending, of course, on what we mean by "intelligence".

"Even a potato in a dark cellar has a certain low cunning about him which serves him in excellent stead."

 

[Bob_the_lost]

I'm with gurrier, that we already have AI. It is very specialised, and it works in the dark, so to speak, without the awareness or creativity that some posters here have explicitly associated with intelligence. But it is all over the place in the modern world. Depending, of course, on what we mean by "intelligence".

Some vocabulary control will likely help the debate along. One widely accepted definition of intelligence is "the faculty to acquire and apply knowledge".

How about: "The faculty of thought and reason."

If what we have now is intelligence then i've got to reconsider about the quality of writing in the Sun.

 

[rich!]

I think it's important to remember that the first time Alan Turing mentioned something like the Test, it was in the context of deciding whether your interlocutor was a man or a woman, using only a Telex machine.

Everyone misses the subtlety of the Turing Test.

Two terminals. You type at both of them. You know the terminals are connected to two different types of thing: man/woman or human/computer or whatever. You have to determine which is which.

Much more interesting than trying to work out if there's a robot at the end of one...

 

[Jonti]

Reason is OK. But to define intelligence in terms of thought strikes me as dangerously circular.

 

[Crispy]

There were no computers until we had a fully formed binary logic figured out. Computers didn't spontaneously emerge from complicated clockwork. Similary, until we have a fully formed theory of mind, true AI will have to wait.

 

[Jonti]

the body is not a linear, clockwork, mechanical machine

Maybe not. But it maintains itself by determinate processes. And we have learned how to interfere with those processes in order both to kill and to cure. As far as those interventions are concerned, the body is an engine whose workings are increasingly well understood.

 

[kyser_soze]

I think the clear division is between AI - which as has been said already exists, albeit in 'insect level' kinds of comprehension - and the fusion of that with identity to create an artificial consciousness - kinda the difference between knowledge and wisdom perhaps?

I remember years ago reading about the computers that are installed to manage baggage handling at O'Hare and Chek Lap Kok airports, and how when they first opened there were loads of problems, misdirected luggage etc because the systems used neural nets to 'learn' the most efficient ways of moving the luggage around - now apparently they are the most efficient in the world. If we consider 'intelligence' to be the recognition of patterns and adapting behaviour etc then really it's just a function of faster processing and better data handling ability. Even creativity is up for grabs - if a computer creates a new solution to a problem off it's own back (for example, the software that was mentioned that's 99.99% right) even tho it's within certain parameters, it's created something.

But intuition - the Eureka! moment of whatever - is a function of consciousness surely? Those moments where ermergence happens, the 'black box' moment, whatever...

Sorry, rambling...

 

[Jonti]

Theory of mind stuff here and here.