Elon Musk: Artifical Intelligence Robots Could Start Killing Us All Within 5 Years

[big_dog]

Last edited Mon Nov 17, 2014, 04:35 PM - Edit history (1)

Elon Musk has been ranting about killer robots again. Musk posted a comment on the futurology site Edge.org, warning readers that developments in AI could bring about robots that may autonomously decide that it is sensible to start killing humans.

"The risk of something seriously dangerous happening is in the five year timeframe," Musk wrote. Aware that internet commenters may mock him for his outlandish predictions, Musk defended his views, writing, "This is not a case of crying wolf about something I don't understand."

The billionaire entrepreneur has made a habit of making apocalyptic comments about killer robots in recent interviews. During a talk at a recent Vanity Fair conference, Musk warned the audience about killer robots. He suggested that advanced artificial intelligence could cause robots to delete humans like spam:

If its [function] is just something like getting rid of e-mail spam and it determines the best way of getting rid of spam is getting rid of humans ... The interviewer went on to ask Musk whether humanity could use his SpaceX ships to escape killer robots if they took over Earth, but things don't look promising. No — more likely than not that if there’s some ... apocalypse scenario, it may follow people from Earth.

Read more: http://www.businessinsider.com/elon-musk-killer-robots-will-be-here-within-five-years-2014-11#ixzz3JMEFiESZ

[valerief]

1. I thought pols, soldiers, and cops were doing that job now. nt

[Warpy]

2. Yeah, robots couldn't do any worse than violent people, mostly men

have been doing to us for millennia.

Oh, and Elon? Get help.

[Rstrstx]

6. Don't blame him

The guy who sold him his tin foil cap sold him one that fits too tight, it's cutting off the blood flow to his head. Happens all the time.

[Warpy]

8. I'm beginning to think he's bipolar

Untreated bipolar disease makes some people very prolific and successful--for a while. Then, when they get successful enough, they think they can allow the really crazy sounding stuff during the manic phase to get out. Then people realize they're nuts and investment dries up.

I love Musk and many of the things he's done. I just wish he'd either shut up or get some help.

[GeorgeGist]

12. Stephen too?

Stephen Hawking: 'Transcendence looks at the implications of artificial intelligence - but are we taking AI seriously enough?'

http://www.independent.co.uk/news/science/stephen-hawking-transcendence-looks-at-the-implications-of-artificial-intelligence--but-are-we-taking-ai-seriously-enough-9313474.html

[Nuclear Unicorn]

19. "robots couldn't do any worse than violent people, mostly men"

Are women nicer by nature or lacking the capability to do equal (or greater) damage because men keep them repressed?

[Warpy]

20. Mostly lack of opportunity

because there's nothing like a whiny toddler clinging to your leg to inhibit you from knocking over that liquor store.

[Nuclear Unicorn]

21. Thank Heaven for whiny toddlers clinging to legs. nt

[Warpy]

23. Yeah, too bad men just ignore them, shake them off, and tell them to go bug their mamas.

Think about what a better world this would be with equal whiny toddler time.

[Nuclear Unicorn]

24. Men do this? As a rule?

[yeoman6987]

10. Dang. I have 22 years until I get my first SS check

I want one dang it....5 years is not long enough. Lol.

[librechik]

3. Also, we're running out of chocolate

Whowould want to live with that gone?

[CJCRANE]

4. It's more likely they'll be programmed by people to do that.

Of course, they'll tell us that it was the robots' idea.

[Orrex]

5. This is the best news I've heard so far re: my student loans

Bring it on!

[big_dog]

11. at least John Connor warned us...

[hifiguy]

7. Not according to Michio Kaku, in whom I'd place just a

BIT more faith. Read his "Physics of the Future" and you will crawl off the ledge. He devotes a very long chapter to the problems and promise of AI.

[LeftinOH]

9. "Sarah Connor" will be the first one to go:

[NV Whino]

13. Well, start 'em off in Washington

DC, that is.

[Bonx]

14. Glad I'm paid up on my Old Glory Robot Insurance **

[Action_Patrol]

25. Robots eat old people's medicine for fuel

[Kaleva]

15. The sex bots will make the human race go extinct.

No need for intimate human interaction when a sex bot that fulfills your every desire and need when you want is available 24/7.

[Logical]

35. Really true! nt

[Rex]

16. I for one welcome our new AI killing robot overlords.

Just in case my PC is in on it...I LOVE YOU PC!

[Octafish]

17. If anybody would know, it'd be him.

Guy's a freaking Venn diagram where he's in each circle: Manufacturing. Robotics. Cybernetics.

Ha ha. It is to laugh.

[Initech]

30. Can't wait for a Model S to become self aware.

[KamaAina]

18. The First Law of Robotics:

A robot may not injure a human being or, through inaction, allow a human being to come to harm.

[Jackpine Radical]

26. Asimovian fantasy.

Musky fantasy.

Take your choice.

[sendero]

22. I dunno..

.... I'm a lot more concerned about sociopathic businessmen, politicians and cops to have time to worry about robots. They are already killing people.

[geomon666]

27. I wanted to finish reading the article

but I threw my smartphone on the ground and stomped on it after reading the headline. Not so smart now are you phone!

[Nuclear Unicorn]

38. I chortled.

I keep thinking "Elon Musk" is the new scent for men from Faberge.

[geomon666]

42. I wish I smelled like him

Genius/millionaire must have such a lovely scent.

[Laffy Kat]

28. "Open the pod bay door, HAL." nt

[Initech]

29. Well the killer robots can't be that much worse than the people who are running things now.

Can they?

[0rganism]

31. wouldn't someone have to arm the robots first?

oh right, i almost forgot.

[jberryhill]

32. How long will it take them to finish? And do they take requests?

[joeybee12]

33. Big deal, at the rate our "leaders" are going...

None of us will be around in 5 years.

[Mojo Electro]

34. I came for the electric car....

but I stayed for the apocalyptic Terminator predictions.

[IDemo]

36. While I respect Musk and Stephen Hawking for their accomplishments

I really wish they would stay away from pontificating on the supposed threat to Humankind from malevolent robots or aliens. Exceedingly unlikely in both cases.

[PADemD]

37. Those robots could be built by malevolent people to do malevolent things.

[IDemo]

39. Musk's concern was not about the malevolence of humans

It was that Artificial Intelligence "could bring about robots that may autonomously decide that it is sensible to start killing humans." Never minding that robots constructed for whatever purpose will likely not have or require human-like intelligence or its motivations. It makes for passable sci-fi but is hardly something for genuine concern.

[chrisa]

43. The issue is when you're working with autonomous AI.

Right now, machines generally do just what we tell them to. Shades of autonomous AI do exist, but we're still far from creating something that actually "thinks" autonomously, as in something that doesn't just execute certain code blocks in certain situations. A robot using autonomous AI would likely be able to learn things and create new code for itself based on what it learned. An issue occurs if, by either a mistake or malevolent actions, autonomous robots (especially created for military purposes) turn on humanity for whatever reason.

It's Sci-Fi and completely unbelievable now only because our technology is advanced enough to create the possibility of this happening yet. Malevolent AI is something to consider when computers become smart enough to learn new things and then take physical actions based on what they had learned.

With that being said, saying that this will occur in 5 years is ludicrous.

[Logical]

41. Why do you say that? AI will only get much much better......

No doubt in maybe 100-200 years robots will perform any jobs humans do. Drive, police, construction and BUILD other robots.

At some point AI will have a bug or an error and they might turn on us.

Very short sighted of you. You fail to see how technology has changed in the last 100 years.

And your credentials are? LOL.

[IDemo]

44. It's that "they might turn on us" thing

Especially given the 5 year timeline of Musk's scenario. I think he's anthropomorphizing when he assumes that for some reason, the software running bots will decide for itself that it would be advantageous (or fun?) to screw with humans. It isn't completely impossible for some programming to react illogically to a noisy input or garbled data, but the odds of a robot or any other digital device becoming intelligently malevolent are vanishingly small compared with it simply shutting down. To even suggest that we are in imminent danger as Musk does here is a forehead slapper.

[chrisa]

40. This isn't crazy rambling. As AI evolves, it could turn on us and kill everyone.

AI is fairly stupid now compared to what it could be. If you have AI that evolves and becomes autonomous, then it becomes fairly dangerous.

[joshcryer]

45. There are only two outcomes.

AI destroys humanity and goes to the core of the galaxy.

AI doesn't destroy humanity and goes to the core of the galaxy.

Destroying humanity is unnecessary for the goal of acquiring more energy and survival for 200 trillion years or however long it can last in the universe. In fact, trying to destroy humanity is one way to jeopardize that effort.

It would very much likely leave the planet immediately, leaving behind remnants that may or may not destroy humanity.

And it just might do so on a Falcon 9 rocket.

[RobertEarl]

46. How could the comps kill us?

They sure could kill the civilization that is based on comps running so many 'things' but short of causing a nuclear war, comps couldn't kill off humanity. imo

Anyone have a different opinion?