Urban75 Home About Offline BrixtonBuzz Contact

Prof Stephen Hawking: thinking machines pose a threat to our very existence

And even then, it can only win at chess under certain conditions. The example I posted above is from a few years ago, and it probably wouldn't fail that particular test now, but I would say that there is bound to still be a position that you could devise that would fool a computer, would expose its lack of understanding about what it is doing.

What understanding is is an important and difficult question, imo, and one whose existence some in AI simply deny - saying that understanding can be computed.

The opposite position, held by Penrose among others, whose arguments to me are very convincing, is that understanding cannot come from any kind of algorithmic computation. There are many examples of the human ability for non-algorithmic understanding that a computer cannot have, and Penrose uses Godel's incompleteness theorem to show how a computer can never have such a thing. We can see the truth behind a so-called 'godel statement'. We can understand it. But a computer can never compute that specific understanding using algorithms, and we can prove that it can't.

I do think quantum computing will be a game-changer, simply because we will be progressing closer to the way we and other life-forms work stuff out - parallel 'try everything, match patterns from other domains' processes that allow for leaps of thought, intuition, inspiration, metaphor.

Show me a computer that understands, and can generate and use, metaphor.
and the one there is if it meets your proof, if its able to write poetry in three languages etc etc. How will you ever know if its real or just a very very clever computer?

Although once you've got a machine that smart you can't tell then it doesn't matter imo. Just accept the HAL. Its convincing.

but it'll trouble the beards and the religious. Doctrine of Souls Revisited: This Time Its Digital
 
and the one there is if it meets your proof, if its able to write poetry in three languages etc etc. How will you ever know if its real or just a very very clever computer?

Although once you've got a machine that smart you can't tell then it doesn't matter imo. Just accept the HAL. Its convincing.

but it'll trouble the beards and the religious. Doctrine of Souls Revisited: This Time Its Digital
I don't rule out real AI on principle. Not at all. I just think we're nowhere near it at the moment, and it's not simply a question of computation and power combined with the appropriate programming. We are not simply universal Turing machines, but that doesn't mean we can't create something that is also not simply a UTM.

I said earlier in the thread, I think, that the first step is to work out how cells work, how worms make decisions. That's how you begin.
 
Where is this going? Pull what plug? We know how to generate electricity now. By various means.

Well, you seem to be saying that if machines get rebellious, we can just "pull the plug" on them and carry on our merry way. I'm trying to point out that it won't be as easy as that. Current civilisation is already heavily dependent on the mass production and distribution of electrical power. It seems reasonable to suppose that will be even more the case by the time that intelligent machines pose a threat to human supremacy. So if we were truly united as a species in attaining the goal of definitively foiling a cybernetic revolt forever, that would require the wholesale destruction of electrical grids and any automated systems we have built up in the meantime (which seems strongly likely in the event that we develop true AI), because otherwise there is the risk that AI could hole up somewhere inconspicuous or otherwise parasitise our energy distribution networks and manufacturing facilities. Oh, and we'd also have to destroy all our computer networks and the physical basis for the internet, because that would certainly be a vector for AI. Actually, even networks wouldn't strictly be necessary if AIs can manipulate events by copying themselves or code written by them onto USB sticks or whatever that humans would then (unwittingly) carry around.

All of this assumes that the entire world can be brought onto the same page regarding how to deal with the threat, or even that the threat exists in the first place. Intelligent machines won't have to be genocidal or even particularly tyrannical for them to maneuver themselves into a position of power or influence. If they can earn the cooperation of a significant portion of humans to enact their rule, whether through persuasion or coercion or both (a good strategy might be to persuade the human elites and coerce the rest, or perhaps vice versa), then that will be enough to end human supremacy.
 
I know about electrical maintenance as I do it for a living. I'm not expecting computers to be able to do it any time soon. It isn't production line work.
I just posted on the basic income thread abour robot brickies. Would have seemed unthinkable a few years ago.

Your job is not safe, no matter what it is.
 
Although once you've got a machine that smart you can't tell then it doesn't matter imo.
Of course, you could create a machine that on many levels is rather thick, but that could also be said to really have a mind. That's another argument I find unconvincing - the idea that intelligence, real understanding intelligence, is a function of complexity.
 
Also, the arguments on this thread are a bit far fetched. It's not going to be robots enslaving humans, like in the Matrix.

One way it could happen is a takeover of the Internet. Imagine a machine with the capability to intercept/attack/change at will everything passing through our networks.

They could 'control' humans indirectly, by changing the information we get from our existing sources. It's already been shown that Facebook can manipulation people into being depressed by changing which posts appear on their wall. Imagine that, scaled up, and across all communication channels.
 
Well, you seem to be saying that if machines get rebellious, we can just "pull the plug" on them and carry on our merry way. I'm trying to point out that it won't be as easy as that. Current civilisation is already heavily dependent on the mass production and distribution of electrical power. It seems reasonable to suppose that will be even more the case by the time that intelligent machines pose a threat to human supremacy. So if we were truly united as a species in attaining the goal of definitively foiling a cybernetic revolt forever, that would require the wholesale destruction of electrical grids and any automated systems we have built up in the meantime (which seems strongly likely in the event that we develop true AI), because otherwise there is the risk that AI could hole up somewhere inconspicuous or otherwise parasitise our energy distribution networks and manufacturing facilities. Oh, and we'd also have to destroy all our computer networks and the physical basis for the internet, because that would certainly be a vector for AI. Actually, even networks wouldn't strictly be necessary if AIs can manipulate events by copying themselves or code written by them onto USB sticks or whatever that humans would then (unwittingly) carry around.

All of this assumes that the entire world can be brought onto the same page regarding how to deal with the threat, or even that the threat exists in the first place. Intelligent machines won't have to be genocidal or even particularly tyrannical for them to maneuver themselves into a position of power or influence. If they can earn the cooperation of a significant portion of humans to enact their rule, whether through persuasion or coercion or both (a good strategy might be to persuade the human elites and coerce the rest, or perhaps vice versa), then that will be enough to end human supremacy.

I said we can exist without electricity and computers can't. I didn't say we can exist exactly like we are now without electricity so it's a moot point really.
 
I said we can exist without electricity and computers can't. I didn't say we can exist exactly like we are now without electricity so it's a moot point really.

Then in that case the question becomes one of how many humans will be willing to return to pre-electrical conditions, with all the deaths and decreases in quality of life that such a move would necessarily entail. I predict that a majority would say "nah, fuck that" and try other ways of dealing with the problem.
 
Brick laying is production line work. It's skilled, but a repetitive sequence.
Fair enough, electricians are more robot-proof than brickies, but I could easily imagine new installations being done by robots. And if installation work is taken out, you'll have a lot more competing for your condition report jobs.
 
Then in that case the question becomes one of how many humans will be willing to return to pre-electrical conditions, with all the deaths and decreases in quality of life that such a move would necessarily entail. I predict that a majority would say "nah, fuck that" and try other ways of dealing with the problem.

Nice straw man. You only need to isolate the supply long enough to decommission this super computer you're fantasising about. Not forever.
 
and the one there is if it meets your proof, if its able to write poetry in three languages etc etc. How will you ever know if its real or just a very very clever computer?l
Sorry, keep quoting this, but it's still relevant to this.

Strikes me that a flaw in the Turing test is exposed by the way we interact on here. In effect we have exactly the conditions of a Turing test now on message boards. And the limitations of the medium are obvious, I think - that's why we need emoticons and why we get so many misunderstandings. We are not just typewriter operators, word generators - we are embodied beings.

And that very embodiedment - there being an 'us' and a 'not-us' - is fundamental to the way in which we develop intention. It's fundamental to the way in which we construct consciousness - 'me in the world' - so to take away that aspect from the test seems to me to be rather unfair, and bound to lead to poor judgements, just as people make poor judgements about the characters of others on here.
 
Fair enough, electricians are more robot-proof than brickies, but I could easily imagine new installations being done by robots. And if installation work is taken out, you'll have a lot more competing for your condition report jobs.

There's nothing stopping electrical services being laid like bricks, preinstalled in the blocks for example. Lights need constant attention though. Maybe computers won't be bothered about lights. :hmm: Actually they'd only need to maintain their supply. That'd be a cunt for them to do though. There's so much involved in it from generation to consumption.
 
Nice straw man. You only need to isolate the supply long enough to decommission this super computer you're fantasising about. Not forever.

Why do you assume that the threat of a cybernetic revolt would only come from one easily-isolated "super computer", rather than being distributed over the internet and various other networks? Even if starts out in one place, why could it not spread copies or child processes of itself or whatever before making a move that humans would unambiguously interpret as hostile?
 
Also, the arguments on this thread are a bit far fetched. It's not going to be robots enslaving humans, like in the Matrix.

One way it could happen is a takeover of the Internet. Imagine a machine with the capability to intercept/attack/change at will everything passing through our networks.

They could 'control' humans indirectly, by changing the information we get from our existing sources. It's already been shown that Facebook can manipulation people into being depressed by changing which posts appear on their wall. Imagine that, scaled up, and across all communication channels.
technically, thats not what happened. They farmed us as meat batteries (because that makes sense in sci fi world) and provided us with a full spectrum sennsorium that mimicked the latter part of the 20th century. Until some ungrateful pricks fucked it all up for everbody and then had a shit rave in their squatted underground shithole.
 
Why do you assume that the threat of a cybernetic revolt would only come from one easily-isolated "super computer", rather than being distributed over the internet and various other networks? Even if starts out in one place, why could it not spread copies or child processes of itself or whatever before making a move that humans would unambiguously interpret as hostile?

Where have I stated I assume that? Cutting electricity would cut all those things. This just reminds me of Superman 3. You can kill a computer pretty quickly. Seconds in fact. With nothing more than water.
 
technically, thats not what happened. They farmed us as meat batteries (because that makes sense in sci fi world) and provided us with a full spectrum sennsorium that mimicked the latter part of the 20th century. Until some ungrateful pricks fucked it all up for everbody and then had a shit rave in their squatted underground shithole.
If that's not slavery....?
 
In fact, straight from the mouth of Morpheus himself:
What is the Matrix? "The Matrix is everywhere...It is the world that has been pulled over your eyes to blind you from the truth." What truth? "That you're a slave Neo. Like everyone else, you were born into bondage, kept inside a prison that you cannot smell, taste, or touch. A prison for your mind."
 
If that's not slavery....?
you could have lived and died a full life indistinguishable from the real where in the real you are in a tank being used for battery power. If you are 'born' and 'die' in a sensorium, a 100% you cannot tell the difference enviroment- well, freedom was I free? Remember the evil traitor in the Matrix trades away reality in exchange for a perfect rendition of a far better time.
 
Where have I stated I assume that?

Because you mentioned a "super computer", which suggests a specialised system in a singular location.

Cutting electricity would cut all those things.

Only for as long as the electricity is switched off. You're aware of course that computers still work just fine when you switch them back on.

This just reminds me of Superman 3. You can kill a computer pretty quickly. Seconds in fact. With nothing more than water.

But as I have repeatedly pointed out, if the threat is distributed over the internet you would have to trash pretty much every network-capable device on the planet, in order to be completely sure that the threat has been eliminated.
 
Because you mentioned a "super computer", which suggests a specialised system in a singular location.

No it doesn't.

Only for as long as the electricity is switched off. You're aware of course that computers still work just fine when you switch them back on.

Are you actually reading my replies?

But as I have repeatedly pointed out, if the threat is distributed over the internet you would have to trash pretty much every network-capable device on the planet, in order to be completely sure that the threat has been eliminated.

You're pulling this stuff out of your arse mate.
 
you could have lived and died a full life indistinguishable from the real where in the real you are in a tank being used for battery power. If you are 'born' and 'die' in a sensorium, a 100% you cannot tell the difference enviroment- well, freedom was I free? Remember the evil traitor in the Matrix trades away reality in exchange for a perfect rendition of a far better time.
Where you know or care if you are a slave or not chances nothing.

It's a similar situation in Meereen in GoT when the freed slaved come to Dany and beg to be allowed to sell themselves back into slavery because their life was better.

They would be there out of choice, but they'd still be slaves. If any of the conditions changes for either the traitor in Matrix or the re-slaved fella in Meereen, they would have no option to remove themselves from their situation. They're slaves.
 
No it doesn't.

What then? How is this notional intelligent machine structured? What is its relation to the internet and other networks?

Are you actually reading my replies?

Are you? I've pointed out various reasons why "cut the electricity" is at best an overly-simplistic solution, but you insist on repeating it as if that's all that was necessary.

You're pulling this stuff out of your arse mate.

What makes you say that? Do you think AI would incapable of copying itself?
 
Where you know or care if you are a slave or not chances nothing.

It's a similar situation in Meereen in GoT when the freed slaved come to Dany and beg to be allowed to sell themselves back into slavery because their life was better.

They would be there out of choice, but they'd still be slaves. If any of the conditions changes for either the traitor in Matrix or the re-slaved fella in Meereen, they would have no option to remove themselves from their situation. They're slaves.
but in the context not of antiquity style slave/churl etc caste- in the context of your body is born in the tank and your mind is born into the utterly flawless rendition of the earth before it got nuked. You literally never know. And your mind is having a whale of a time, lobster and all sorts- yes the machine holds your flesh in bondage but if the choice is buying Morpheus a a J20 at that shitty zion rave instead then fuck that shit, upload me right now.
 
Back
Top Bottom