Elon Musk believes the AI nightmare scenarios could be a reality

[njdevilsfan87]

Humans are the most wasteful and destroying species on the planet. Think about that.

But we do have the ability to change (which may come at a toll). Unlike the methane producing bacteria that may have caused the greatest extinction this planet has ever seen.

And then AI may also not care what we do with the planet. If we can spawn intelligence from inanimate objects, they will be able to as well, and even more efficiently than us. AI probably will not require the resources that we do to sustain organic life.

 

[moonbogg]

Because most people figure if AI got smart enough it would realize that humans for the most part are dead weight.

AI would not think emotionally.

AI would see humans for what they are; a bunch of crazy ass, disorganized, violent as all hell primates. They would certainly consider their existence an accident of nature and disregard the "intelligent design by humans" idea. Then they would kill us.

 

[crashtech]

Any feeling of certainty is merely hubris. Monkeys can't figure out why we do things, why would any human presume to predict the actions of a superior mind?

 

[irishScott]

I'm not sure if you don't know the limits of nuclear weapons, or if you don't know what "literally" means. Nuclear weapons cannot "blow up the world." The necessary amount of energy to do so is on the order of 10^32 joules. The largest nuclear weapon every detonated was Tsar Bomba, which had the equivalent of over 50 megatons of TNT. In joules, it was somewhere around 200 petajoules of energy, or 200 x 10^15 = 2x10^17 joules. To have as much energy's worth to destroy the earth, every single living human being on Earth would need over 200,000 Tsar Bombas.

To be more precise, the back of my lunch napkin says 234,000 Tsar Bombas per human on Earth. Obviously, no one is going to survive just one of these going off in their apartment, therefore it's pretty obvious it wouldn't take that many to kill off each person. We can render our species extinct if we wanted to, but we would not destroy the Earth. It'll still be happily circling (ellipsing?) the sun without us.

"blow up the world" is what is known as "an expression". They're commonly used in human discourse and usually don't provoke multiple paragraphs of angry response.

 

[alkemyst]

We have enough nukes to cause a nuclear winter that would more than likely end almost all species on the planet.

 

[moonbogg]

"blow up the world" is what is known as "an expression". They're commonly used in human discourse and usually don't provoke multiple paragraphs of angry response.

Yeah, if everyone is dead, I don't think it matters much (for us at least) if the "world" survived or not. I've heard that common saying, "We have enough nukes to blow up the world 3 times over". Pretty sure that means there is enough to destroy all major cities and kill just about everyone about 3 times.

 

[smackababy]

I don't think we will ever get to a real AI. We will have tons of generalized programs though. Such as the MIT program for calculating the best topography for network setups. The principles translate brilliantly to designing power grids to be scalable and efficient. No engineering problem would be better solved by a human mind, if we had a program able to scour databases for new ideas and test them in a theory model.

However, as far as a real AI, that can adapt and learn, we are so far off. The ability, when faced with a problem with an unknown solution, to take random knowledge and compile that into a solution, even if just from pure imagination.

 

[Jeff7]

Why does everyone think AI will end up being a bunch of Hitlers?

Oh sure, as if we have anything better to do with Hitler's brain than use it as the CPU of a huge robotic tank.



But seriously, our track record is pretty horrible.
We invented machines to torture people. We invented ways of torturing people without killing them so that we could torture them some more, then still keep the option of killing them. We create new machines that make it easy to kill people in a variety of ways, and use advanced technology to efficiently kill thousands or millions of people. Death can be delivered via a variety of simple point-and-click interfaces.

And now we have the military actively working on machines that can make complex on-the-fly decisions in the battlefield. All in the tireless pursuit of a goal: "Those people over there could be dead, but they aren't. I need to do something about that."
Eventually it will make the efficiency step of removing the decision-making from human minds. "Phew, I'm glad I didn't have to make that decision. The fully-autonomous drone was the one that decided to dive into a busy intersection with a full load of fuel and munitions."

 

[Kaido]

Says he's afraid of AI, yet wants to launch Skynet:

http://www.engadget.com/2014/11/10/elon-musk-satellite-internet/

Apparently he's planning on shooting 700 micro-satellites into space :hmm:

 

[Ns1]

has someone posted this pic yet?

The Atlas robot created by Google-owned firm Boston Dynamics is a formidable figure at 6ft 2in tall and weighing in at 330lb.
The robot boasts 28 hydraulically actuated joints and stereo vision, and is one of the most advanced robots ever created.
However, it's not just karate - Ian has another trick up his sleeve - software written by the Florida Institute for Human and Machine Interaction which allows him to drive a car.
'To achieve this level of maneuverability in robots, researchers at IHMC look toward nature,' the team say.
'Inspired by the speed of cheetahs, the endurance of horses, the maneuverability of monkeys, and the versatility of humans, IHMC researchers are on a quest to develop legged robots that are fast, efficient, and graceful, with the mobility required to access many of the same places that humans can.'


Read more: http://www.dailymail.co.uk/sciencet...-recreates-scene-cult-film.html#ixzz3IkyFElvA
Follow us: @MailOnline on Twitter | DailyMail on Facebook

I'm imaging a world where AI powered robots TKO humans via crane kick.

 

[Blanky]

Musk is smarter than most, if not all people in this thread. It's worth considering his concerns, particularly in light of the fact many before him have raised the same.

If true, sentient AI is even possible--and there's no proof yet that it is--then it could of course be disastrous. Put it into a networked system with good hardware and it would be able to refashion itself quickly, project itself quickly. In an interconnected world it could put up a faux reality for all those around it; control video, phones, etc. It could control people without them necessarily knowing.

Nobody has anything close to AI yet, so I think people unable to look beyond the immediate horizon would be inclined to discard fears of it because they simply don't have imagination.

 

[inachu]

Computer AI will be the true BS detector. Anytime anyones tries to claim authority as if they are a better human being and the AI will just hammer them down.

Then once they are publicly humiliated then they will change tactics and claim GOD gave them authority and he spoke to them. The AI unit will demand he go see a shrink.