Elon Musk: 'We are summoning the demon' with artificial intelligence

cnet.com ^ | October 26, 2014 10:09 AM PDT | Eric Mack

[BenLurkin]

"You know all those stories where there's the guy with the pentagram and the holy water and he's like... yeah, he's sure he can control the demon, [but] it doesn't work out."

This has become a recurring theme in Musk's public comments, and each time he warns of the AI bogeyman it seems even more dire.

In June, Musk raised the specter of the "Terminator" franchise, saying that he invests in companies working on artificial intelligence just to be able to keep an eye on the technology. In August, he reiterated his concerns in a tweet, writing that AI is "potentially more dangerous than nukes." Just a few weeks ago, Musk half-joked on a different stage that a future AI system tasked with eliminating spam might decide that the best way to accomplish this task is to eliminate humans.

[JohnBrowdie]

someone has been watching too many movies.

[Obadiah]

The more I hear from this guy the more I think he's a crackpot. But he has all of liberaldom believing with certainty that he is a savant genius.

[grania]

(read later)

[BenLurkin]

Yeah...what does a guy named “Musk” know about anything anyway....

/s

[pepsi_junkie]

Just a few weeks ago, Musk half-joked on a different stage that a future AI system tasked with eliminating spam might decide that the best way to accomplish this task is to eliminate humans.

Then they have a programming fault. Spammers aren't human.

[Da Coyote]

Musk appears to be a shrewd businessman.

However, much like Bill Gates, he apparently has the tech savvy of the Obamadork.

[quantumman]

He is a hopeless reductionist. Machines cannot be made to think per Turing’s Halting Problem. Given a task with no answer (This statement is false - determine its provability) a machine cannot determine its absurdity. Kurt Godel opined and proved this back in the late ‘20’s with his Incompleteness Theorem.

[shove_it]

AI’s been coming on stronger lately and they’re not just driving your car.

“Humans Need Not Apply” ...

https://www.youtube.com/watch?v=7Pq-S557XQU

[DannyTN]

The problem isn't AI by itself, but a sufficiently enabled AI that has access to sufficient robotics. Robotics without AI is relatively benign. They do exactly what you program them to do and nothing more. I said relatively, because they can be programmed to kill, and you don't need AI for that. AI by itself is completely benign. An intelligence trapped in a computer not hooked to the internet, can at best only advise or influence it's human interactors. An AI that depends on man for it's maintenance and/or power supplies is weak. An AI that depends on robotics but still depends on man for supplies to build the robotics is still weak. An AI that has sufficient robotics that can obtain it's own resources, is a potential threat. Self replicating nanobots would be one such scenario. A lot depends on how the AI's higher order thought processes are constructed. Whether it has a value system. Whether it has prime directives. The potential for unintended consequences with AI is hugh! :) Imagine:

• Tasked with reducing man made pollution, it chooses to eliminate the source.
• Tasked with reducing abortion, it sterilyzes man.
• Tasked with reducing global warming (and fed fake data to believe it's real), it triggers volcanic eruptions and sends us into an ice age.
• Tasked with protecting us from Ebola, it puts in travel restrictions. (sarcasm)
• etc.

[MrB]

I’m not afraid of AI as it will never reach the point of being dangerous.
I base this on my own worldview assumptions about the nature of reality.

[precisionshootist]

I think he should be more worried that AI might just reason that liberals are the problem.

[Axenolith]

I used to consider this, then I realized that anything we create that becomes “self aware” will have the same foibles and desires as any other self aware organism.

It won’t be able to use it’s impeccable super mind to it’s fullest because it’s going to be worried about how it appears to others, how is it going to pay the food (power) bill, “do these cooling vanes make me look fat?”, “Is A9765 trying to snipe my promotion”, “Are humans gods?” “Is their god my god?” etc...

[Moonman62]

Computers can get cranky when they are conflicted.

[Moonman62]

I used to consider this, then I realized that anything we create that becomes “self aware” will have the same foibles and desires as any other self aware organism.

...

Eaxactly, and such advanced machines will certainly have a “fugetaboutit” routine.

[sonofagun]

An intelligence trapped in a computer not hooked to the internet, can at best only advise or influence it's human interactors.

What if the Internet IS the AI?

I don't beleive one machine, even a supercomputer will have the resources to surpass human intelligence.

However, billions of networked PC's, Tablets, smartphones, DVR's, etc. may very well surpass us in ways we cannot even imagine right now.

[xp38]

The current season of Person of Interest has two supercomputers in a war against each other. One is the machine that Harold Finch designed from the start of the series and is supposed to be the good one and a usurper evil computer called samaritan. Should be a fun season watching them slug it out using human pawns to do their bidding.

[Bobalu]

But then again the creators of an advanced AI might advise it against reading “On computable numbers, with an application to the Entscheidungsproblem” and admonish it to love Hilbert and eschew Gödel. ;-)

And in its arrogance it might proceed to take over the planet while ignoring its own ignorance and the impossibility of its being.

[DannyTN]

"Machines cannot be made to think per Turing’s Halting Problem. Given a task with no answer (This statement is false - determine its provability) a machine cannot determine its absurdity. Kurt Godel opined and proved this back in the late ‘20’s with his Incompleteness Theorem."

An AI can be programmed not to spend an inordinate amount of time on any one problem. In which case it wouldn't matter if it was given a problem that was absurd, it would hit the time limit and set it aside. Heck windows basically does that now when a task doesn't respond.

An AI can be programmed to recognize absurdity. It wouldn't recognize every situation of absurdity, but then neither do we. We struggle with problems that are eventally proved to be absurd.

[Harmless Teddy Bear]

This kind of stuff reminds me of Y2K when people were telling me that the computers would think it was 1900 and since planes had not been invented then all the planes would crash.

Only a human could come up with logic like that.

Computers are very big filing cabinets with a very fast retrieval system. If you think that is all "intelligence" is then I suppose "AI" is possible.

But it is far more then that.