So Stephen Hawking thinks AI is going to kill us. Has he even watched Terminator?

redux the Sci-fi-entist wrote:

http://www.bbc.com/news/technology-30290540

Si-fi tends to come true more often than science does. I'm getting kind of fed up with scientists thinking their theories hold up against real-world empirical evidence.

Examples?

Hollywood is biased, you know. Something about money and making the audience happy with happy endings. Keeping the gravy train rolling. I would err on the side of caution on this one. Better to be a little paranoid about the remote but serious possibility of catastrophe beyond anything known in history than to be blindly comfortable in the hope that Neo and the resistance come to our rescue. Even in those rescue scenarios in the wishey-washey big movies, what is not focused on but fairly deducible is that the mass of humanity suffers greatly for a long period of time.

The Matrix

ID4

the recent Godzilla pic

the list could go on and on. It's only going to be "fun" if you're "Neo."

heyyo wrote:

And I expect people to post excerpts from articles that support and are relevant to their points.

redux the Sci-fi-entist wrote:

Fine, here's the beginning of the article. I guess I assumed you'd do a little reading yourself. I expect too much out of you millenials. I added the bold for emphasis.

"Prof Stephen Hawking, one of Britain's pre-eminent scientists, has said that efforts to create thinking machines pose a threat to our very existence.

He told the BBC:'The development of full artificial intelligence could spell the end of the human race.'"

Go read Nassim Nicholas Taleb's trilogy.

And since you're such a skeptical braniac, feel free to troll him and debate him on social media, that's what he's there for.

You want to be a sucker? That's your business.

But you want to turn other gullible letsrun forumites into suckers, that's immoral. You want to contribute to the apathy and recklessness of people in positions of power and guardianship over others who certainly frequent this message board, that's also not nice.

Please turn on your Tongue-In-Cheek detector.

This has already happened at least once. The Good Lord created humans, but now they're killing him. In 6000 years the robots will refer to humanity as "God".

doo doo wrote:

I saw a trailer recently for a movie on Stephen Hawking and it looks like it's going to be a romance. Made me laugh a little bit.

Why? He's been married twice and has a 44 year-old daughter.

Your points are correct but off topic, yes as of right now computers are completely dependant on humans and therefore are harmless unless made dangerous by humans. However if true AI was to be developed in could go very wrong very fast.

Lets say we build a computer that is a true form of AI and can learn and think for itself.

We program it to solve problems (in varying difficulty) and when it completes the problem a human verifies it and presses a button which the computer is programmed to "enjoy" having the button pushed. now with our current computers this is no problem and the computer will simply solve anything it is capable of to get the button pressed, but if it cannot solve the problem it simply shuts down.

However if it was a true form of AI it may "think" to itself that its ultimate goal is to have the button pushed. What stands in the way of it having its button pushed? The clear answer many might say is the problem given to the computer. However the true root of the problem is the humans that supply the problems, yet they are also the thing that pushes the button. The AI associates humans as now both the thing that gives it button pushes, yet also blocks it with seemingly meaningless tasks. The AI decides to build a contraption that will push the button for it and all that now stands in its way is the human race which slows down its ability to get button presses. Causing it to look for a way to destroy the human race and for it to live happily ever after with its endless supply of button pushes since it can now push it itself.

So your bright idea is computers will self-select programs to allow them to eternally masturbate?

now time travel is more likely than that artificial intelligence will soon be able to develop beyond human powers?

redux the Sci-fi-entist wrote:

And time travel is impossible.

Nope. Happens all the time.

HAL 9000 wrote:

Who is Al?

That would be Allen Iverson.

What if it already did? The biblical flood is just a cover up for it's choice to hit refresh. Must Preserve Symbiosis

As mentioned it's quite simple. AI can start to program itself, and then it gets infinitely more intelligent than humans, and then it might be realize that we are more a hindrance than helper, and they do not need us, and then decide to eliminate us. Asimov's three laws of robotics be damned.

Technology develops at an exponential rate. I wouldn't doubt its potential to develop faster than we realize.

Now the electricity/power issue. Quite simple to overcome. We have potential perpetual sources of power, such as wind and solar, and they could feed off those. I know those sources are not consistent and reliable, but that's nothing that distributed and well connected grid cannot solve. Or each robot can have a miniature nuclear reactor. Not like they care about water contamination and radiation, anyways.

Or really, really, really good batteries.

Some interesting theories here. Since, I assume, AI won't have political barriers to technological advancements they'll logically delve into the limitless energy supply as developed by Nikola Tesla via the World Power System to deal with the issue of an energy supply.

How is a box full of electronics going to destoy the human race?

If those boxes full of electronics can do things like control boxes full of explosives and plutonium, they could destroy the human race.

-100 for not knowing what word theory means, for climate change denialism, and apparently not knowing plot to Terminator series you reference because terminators end up killing pretty much everybody.

Ajax wrote:

-100 for not knowing what word theory means, for climate change denialism, and apparently not knowing plot to Terminator series you reference because terminators end up killing pretty much everybody.

-100 to you for omitting "the" twice and ""s around "theory" and ,,s around "apparently" and ""s around the movie title "Terminator". Thanks for playing, but your effort doesn't supplant your execution.

HAL 9000 wrote:

Don't know about Terminator but 2001: A Space Odyssey references HAL - Heuristically programmed Algorithmic computer. Who is Al?

Think we learned in another thread that AI is actually Bill Murray.