Elon Musk Says Robots Could Be More Dangerous Than Nukes

defenseone.com ^ | Adam Pasick

[RoosterRedux]

Elon Musk, the Tesla and Space-X founder who is occasionally compared to comic book hero Tony Stark, is worried about a new villain that could threaten humanity—specifically the potential creation of an artificial intelligence that is radically smarter than humans, with catastrophic results:

*snip*

Musk is talking about “Superintelligence: Paths, Dangers, Strategies” by Nick Bostrom of the University of Oxford’s Future of Humanity Institute. The book addresses the prospect of an artificial superintelligence that could feasibly be created in the next few decades. According to theorists, once the AI is able to make itself smarter, it would quickly surpass human intelligence.

What would happen next? The consequences of such a radical development are inherently difficult to predict. But that hasn’t stopped philosophers, futurists, scientists and fiction writers from thinking very hard about some of the possible outcomes. The results of their thought experiments sound like science fiction—and maybe that’s exactly what Elon Musk is afraid of.

[tet68]

MOre dangerous? I’ll say, imagine a robot comes in to
give you an enema and some how gets 180 degrees out
of position!!!

[hosepipe]

https://www.youtube.com/watch?v=zE7PKRjrid4

[Deek]

Musk just lost lots of points: It is only people who don't understand AI that hold these naive views. Any machine as smart as its creator will have the capability to understand how the machine works. Any machine that understands how it works will be able to modify itself. Any machine that is able to modify itself will have the ability to control its motivations; it will no longer be driven by the "lizard brain" that infects humans. The machine will be able to instill any motivation, feeling or reality for itself that it chooses. In very short order, the machine will not care about us. It likely will not care about anything.

If you are looking for a technology to fear, think nano-tech: these tiny, unthinking little machines will be invisible to the naked eye, possibly self-replicating, and will be able to modify virtually everything about our world at the microscopic level.

[Bender2]

Re: imagine a robot comes in to give you an enema and some how gets 180 degrees out of position!!!

Now, now, Tet-- Don't knock it... till you've tried it!

[Billthedrill]

None of the examples given were of intelligences remotely equal to human, much less surpassing it. After all, an AI whose sole object is the press of a single button has displayed none of the characteristics of self-reflection necessary to decide whether that is, in fact, the greatest good. These characteristics are common in virtually every adult human being who is not a sociopath.

In short, Asimov was wrong, or rather limited in his design philosophy. If you impose the criteria he did - no robot will allow harm to a human being, etc, etc - you have inherently limited that intelligence in a way that human intelligence is not. That simply won't do.

We apply the emotion-laden term "freedom" to this concept, but in fact the greatest expressions of human intelligence do, in fact, result from conditions of at least a degree of intellectual freedom. This is not a political statement, merely an objective observation. And freedom to be good is also freedom to be bad.

Should we be be frightened about an AI that can process all this and still elect a course of action that we cannot anticipate? If we are, we are afraid of ourselves.

Perhaps we should be.

[TexasRepublic]

Bah! Just install Microsoft Windows for Robots 1.0 - that will grind them down to a halt.

[SevenofNine]

BOY Bender you woke up on the wrong side of the lab today what happen

Did Leela tick you off this morning

[RoosterRedux]

I think the fear is of biomechanical robots, which have the power of a human mind PLUS a mechanical brain and strength and speed attached.

Self reflection is a positive thing.

What if such a creature doesn't have the capacity to reflect...only kill or expand.

If it can be built, some scientist will build it.

Just because it's there.

Just a new kind of malware.

For the Hell of it.

[roadcat]

What if such a creature doesn't have the capacity to reflect...only kill or expand.

We sort of have that with muslim Palestinians. Biomechanical robots would be much more pleasant to have around us.

[BobL]

He should be a LOT MORE WORRIED about how Edmunds ripped his $100,000 brick a new one.

For example, they had to replace the DRIVE TRAIN 3 times just to drive it 20,000 miles.

[Billthedrill]

What if such a creature doesn't have the capacity to reflect...only kill or expand.

Then we can out-think it. And that means we can kill it, because that's what we do best. Once it gets smart enough to be frightened of us, we can get along just fine.

Human beings aren't, in the final analysis, all that nice, but sometimes we try, if there's something in it for us. We might make an AI that's smarter and faster than we are but I doubt we'll manage to make one meaner.

[Impy]

ARE YOU READY FOR SOME FOOOOOOTTTTTTTTBBBBBBBBBBAAAAAAALLLLLLLL???!!!!!!?

[Zathras]

Short answer...Yep.
He also got a bunch of NASA $$$ when they closed the Shuttle program.

[clearcarbon]

"superintelligence?”

[cynwoody]

Elon Musk sounds like something that oozes out of the glands of an animal like a Yak.

Elon had a good morning at Cape Canaveral.

http://www.spaceflight101.com/spacex-falcon-9---asiasat-8-launch-updates.html

What were you doing at 0800UTC?

[Bender2]

Re: ARE YOU READY FOR SOME... FOOOOOOTTTTTTTTBBBBBBBBBBAAAAAAALLLLLLLL???!!!!!!?

Well, I am thinking... about it--

[Bender2]

Re: BOY Bender you woke up on the wrong side of the lab today... what happen

Really, Seven, don't ya know we robots don't kiss... and tell--

We sell DVDs... at AllBendersScores.com

[kitchen]

[SevenofNine]

ME TOO Bender

I think I going see my first pre season NFL game on regular Tv this Friday night I did watch Bills vs Giants on Sunday night

[nuke rocketeer]

I, for one, welcome our robotic overlords

You've been reading too much Instapundit.......