8 ms·

Elon Musk, Zuckerberg Trade Barbs Over Artificial Intelligence

28 points by SirLJ 9y ago

WheelsAtLarge 9y ago
Advances in tech should not overpower Society's needs. The first thing people do with tech is to use it to create or enhance weapons. Anyone that denies that needs to look back at history and see how technical innovations have been used. It's time to discuss the dangers of tech and how to limit its ability to create massively dangerous weapons. We all need to take the rose color glasses off and deal with the problem.
- nnfy 9y ago
  >The first thing people do with tech is to use it to create or enhance weapons. That's only true if you're cherry picking. Plenty of tech doesn't even apply to the defense industry. Moreover, I'd posit thay any limitation placed on tech to limit danger will almost certainly limit potential benefit as well. Ignorance is never the solution.
  - WheelsAtLarge 9y ago
    No, pick every advancement/breakthrough in technology and you'll see it either enhanced a weapon or created a new class of weapon. No cherry picking. If it can be used in society it can be used in war.
    - nnfy 9y ago
      Cotton gin Hypodermic needle Fracking Solar panels This argument is a little pedantic, but here are some examples that contradict your assertion. Consider this: without the bomb we would not have nuclear power. Without modern warfare we would not have modern medicine. There are other examples, but you would have been worse off limiting these. Not to mention non-technological benefits like geopolitical stability from MAD.
      - trophycase 9y ago
        Well I think what OP actually means (how I understand it) rather than specific inventions is "meta-technologies" such as AI, cryptography, the internet, etc. Perhaps meta-technology isn't the right term or super clearly defined but hopefully you can grok what I mean. Clearly not every single invention was created for he military.
      - WheelsAtLarge 9y ago
        > Cotton gin Better clothing for soldiers, tents, and more clothing for them. Plenty more I'm sure. Early war planes used canvas instead of metal as outside cover as they went out to fight each other. > Hypodermic needle Enhancement of battlefield medicine. Fix them faster so they can continue to fight. > Fracking Cheaper and continued source of fuel for war machinery > Solar Spy satellites, power for field communications. It's only the beginning. Nuclear power is a perfect example of the dangers of technology. The world powers have decided to limit its spread because it's so dangerous to let everyone have it. If nuclear bomb tech was free to run its course they would be small, cheap and every country would have some. And most would not hesitate to use them.
        dsacco 9y ago
        "Making clothes for soldiers" is reaching. You're expanding "use technology for war" to the point of nonsense and diluting the point. I can just as easily say that the cotton gin enhanced every thing ever because it allowed every human involved in every endeavor to wear clothes. Furthermore, to circle back to your original point: even if we accept that the cotton gin is a technology "used for war", it most certainly was not the first nor the primary use of that technology (the first use was clothing).
        Veedrac 9y ago
        You're cherry-picking an irrelevant property of the argument. So what if, hypothetically, general AI is used nth rather than 1st in line for warfare? It's still AI used for warfare.
        9y ago
    - TeMPOraL 9y ago
      That's tautological. The very natures of technology and war mean that any and all technologies can, and will, be used for military, directly or indirectly. As you say, "If it can be used in society it can be used in war". But this is not a good argument against technology or research.
shusson 9y ago
As Rodney Brooks was recently quoted: "There are quite a few people out there who’ve said that AI is an existential threat: Stephen Hawking, astronomer Royal Martin Rees, who has written a book about it, and they share a common thread, in that: they don’t work in AI themselves" [1] [1] https://techcrunch.com/2017/07/19/this-famous-roboticist-doesnt-think-elon-musk-understands-ai/ https://techcrunch.com/2017/07/19/this-famous-roboticist-doe...
- ericb 9y ago
  Elon Musk works in AI--he's leading an attempt to make truly autonomous vehicles.
  - shusson 9y ago
    Yeah it's a loose definition, but I think the point is Musk does not develop or research AI directly. He owns a company that is developing AI.
- Aron 9y ago
  That's gotta be some kind of pick and choose fallacy (availability bias?). Those two happen to be high-status individuals from old high-status science disciplines so they tend to get whatever they say amplified. I wouldn't use this kind of reasoning to figure out what opinion AI professionals as a class have, but I do know there are more celebrities outside of AI than inside of AI.
- WheelsAtLarge 9y ago
  It doesn't take much AI to create smart weapons. A smart drone can do a lot of damage. I don't even want to think about a level 5, completely automatic, car, which is what tech visionaries are aiming to create. Tech creators look at the bright side of their creations. They have tunnel vision and can't foresee the harm. If they can't others need to. The criticism that Musk does not know enough about AI ignores the fact that he is trying to create a fully automatic car with AI and has experience with its limitations. Yes AI(a catch all process for making computers do things that require intelligence when done by humans) is a process and it has no intelligence in itself but you can say the same thing about C++. It's just a language but look at what's possible to do with it.
hb3b 9y ago
Elon's tweet: https://twitter.com/elonmusk/status/889743782387761152 https://twitter.com/elonmusk/status/889743782387761152
- dsacco 9y ago
  I'm normally a huge fan of Elon Musk, but that tweet comes across as pretty hostile to me. Neither party seems to have critically backed up their beliefs, but I didn't get the sense that Zuckerberg was attempting to single out any individual for a lack of understanding.
ilaksh 9y ago
I don't think people like Musk or the others that he gets some of these ideas from, such as Kurzweil, are actually saying we shouldn't pursue advanced AI. Are they? I mean Musk helped fund OpenAI right? He is just saying that it is actually an existential threat or will be soon enough that we need to consider the problem seriously. I think the point of him mentioning that besides just worrying about it, is to motivate efforts like Neuralink that will allow for high-bandwidth communication between the brain and AIs. There is some misunderstanding on the part of Zuckerberg and others that saying that AI is an existential risk is the same thing as saying 'we should suppress AI research'. You would not think that was the implication if you understood from the perspective of Kurzweil/Musk/other people who have read Kurzweil's/Bostrom's book(s) or think the same. The idea is that we cannot possibly stop advanced human-like and then human-surpassing AI from developing. No matter how hard we try to slow down AI development. So when they say that AI can be dangerous, they do not mean that we should try to stop developing AI, since that is impossible. What they mean is that we should try to mitigate the risks to some degree. One way, which Musk and others are pursuing, is to try to create strong but friendly AI before some less friendly but equally potent variant becomes dominant. But if we do not develop some kind of high bandwidth interface with computers and AI that we control, eventually we should expect computer-based intelligence to significantly surpass our capabilities, to the point where we become irrelevant. If you disagree that there is any possibility of strong AI or any existential threat, what harm is there in pursuing strong but friendly AI, or in developing the BCIs (brain-computer interfaces)? The only 'harm' I can see is if this gets oversimplified as it has been to the idea of "AIs could be bad, we must stop AI" -- which they are not trying to say.
- bredren 9y ago
  Wasn't Elon's original point that he has current access to working AI demonstrations that personally frightened him, and served as the basis of his call for regulation?
  - tim333 9y ago
    Kinda though his arguments seem largely from first principles that robots will get smarter than us so we'd better be careful what they get up to. Here's him chatting a bit at "National Governors Association Streamed live on Jul 15, 2017" https://www.youtube.com/watch?v=2C-A797y8dA&feature=youtu.be&t=48m9s https://www.youtube.com/watch?v=2C-A797y8dA&feature=youtu.be...
ExactoKnight 9y ago
No, it's Elon Musk whose understanding of AI is limited. I doubt Musk has ever actually built a neural network, or got close to coding one... because if he did, he would realize just how fragile and hard to generalize they are. Stacking a ton of data together to build predictive models that can produce intelligent-like outputs, is a far, far cry from a robot gaining consciousness or a sense of agency.
- r_singh 9y ago
  With Musk talking about AI the way he is, seems to me that he wants to use his reputation and the public's naivety to instill fear and have some control over the distribution of AI that works for general purposes...sounds like a great opportunity. If he stays consistent with what he says about AI, I think he might be successful in deterring people from using useful and reasonable AI for a while. I wish he would take the dangers of new features like Autopilot in a production car as seriously though.
  - TeMPOraL 9y ago
    > If he stays consistent with what he says about AI, I think he might be successful in deterring people from using useful and reasonable AI for a while. What exactly did he say that would imply that? As far as I know (and I did watch all the recent interviews), his approach boils down to: - more open access to AI is less dangerous than it being developed in secret by corporations or governments; hence OpenAI - suggesting that regulators start paying attention (and explicitly advising them not to start regulating outright; just to form an agency that would "gain insight" into the field)
    - ExactoKnight 9y ago
      I do agree with this. One of the clearest inequalities emerging in A.I.-driven machine learning is how useless those techniques are unless you have access to a monstrously large amount of well categorized data. Currently you need to be one of the big four to have that access, and this is highly monopolistic.
cartyboi29 9y ago
My nigga Elon gonna pull up to zuckerfucks house with a 44
mncharity 9y ago
For a bit more thoughtful and nuanced discussion, the recent AI Now 2017 - Public Symposium[2] stream[1] (2.5hr) wasn't bad (at 1.5x speed). [1] https://www.youtube.com/watch?v=npL_UsK_npE&t=285 https://www.youtube.com/watch?v=npL_UsK_npE&t=285 [2] https://artificialintelligencenow.com/schedule/2017-symposium https://artificialintelligencenow.com/schedule/2017-symposiu...
richev 9y ago
Sounds like Musk has been reading the 1997 book written by my old Cybernetics professor Kevin Warwick (aka Captain Cyborg)... https://en.m.wikipedia.org/wiki/March_of_the_Machines https://en.m.wikipedia.org/wiki/March_of_the_Machines
bryanrasmussen 9y ago
That article just seemed to run out at the end, it seemed like there was some previous problem between the two related to the Shuttle? And then the article didn't describe it where I was reading.
nradov 9y ago
This debate seems so premature. It's hard to take seriously when so far no one has even come close to building an AGI equivalent to a mouse, let alone a human. Here is a broader set of perspectives from people working in the field. http://spectrum.ieee.org/computing/software/humanlevel-ai-is-right-around-the-corner-or-hundreds-of-years-away http://spectrum.ieee.org/computing/software/humanlevel-ai-is...
- _archon_ 9y ago
  If we have an AGI equivalent to a mouse, wouldn't it iterate itself to a human intelligence pretty rapidly? Safeguards and intelligent thinking now could be the difference between: Mouse -> cat -> shark -> Evil human mastermind -> problems and: Mouse -> cat -> dolphin -> helpful friendly human -> better world The end situation is highly theoretical, but a self-modifying AGI will think differently than a human does, so it's hard to predict what could happen with our existing preconditions. With such a broad range of potential divergence, and a steep difference in desirability of potential outcomes, it seems reasonable to ask some pointed questions sooner rather than later.
  - nradov 9y ago
    No there's no evidence to indicate that would happen. Mice have been around for a long time. How much smarter is a mouse today relative to a mouse a million years ago? And why would a simple mouse level AGI have the intelligence to improve its own abilities?

28 points by SirLJ 9y ago

Advances in tech should not overpower Society's needs. The first thing people do with tech is to use it to create or enhance weapons. Anyone that denies that needs to look back at history and see how technical innovations have been used. It's time to discuss the dangers of tech and how to limit its ability to create massively dangerous weapons. We all need to take the rose color glasses off and deal with the problem.

nnfy 9y ago
>The first thing people do with tech is to use it to create or enhance weapons. That's only true if you're cherry picking. Plenty of tech doesn't even apply to the defense industry. Moreover, I'd posit thay any limitation placed on tech to limit danger will almost certainly limit potential benefit as well. Ignorance is never the solution.
- WheelsAtLarge 9y ago
  No, pick every advancement/breakthrough in technology and you'll see it either enhanced a weapon or created a new class of weapon. No cherry picking. If it can be used in society it can be used in war.
  - nnfy 9y ago
    Cotton gin Hypodermic needle Fracking Solar panels This argument is a little pedantic, but here are some examples that contradict your assertion. Consider this: without the bomb we would not have nuclear power. Without modern warfare we would not have modern medicine. There are other examples, but you would have been worse off limiting these. Not to mention non-technological benefits like geopolitical stability from MAD.
    - trophycase 9y ago
      Well I think what OP actually means (how I understand it) rather than specific inventions is "meta-technologies" such as AI, cryptography, the internet, etc. Perhaps meta-technology isn't the right term or super clearly defined but hopefully you can grok what I mean. Clearly not every single invention was created for he military.
    - WheelsAtLarge 9y ago
      > Cotton gin Better clothing for soldiers, tents, and more clothing for them. Plenty more I'm sure. Early war planes used canvas instead of metal as outside cover as they went out to fight each other. > Hypodermic needle Enhancement of battlefield medicine. Fix them faster so they can continue to fight. > Fracking Cheaper and continued source of fuel for war machinery > Solar Spy satellites, power for field communications. It's only the beginning. Nuclear power is a perfect example of the dangers of technology. The world powers have decided to limit its spread because it's so dangerous to let everyone have it. If nuclear bomb tech was free to run its course they would be small, cheap and every country would have some. And most would not hesitate to use them.
      - dsacco 9y ago
        "Making clothes for soldiers" is reaching. You're expanding "use technology for war" to the point of nonsense and diluting the point. I can just as easily say that the cotton gin enhanced every thing ever because it allowed every human involved in every endeavor to wear clothes. Furthermore, to circle back to your original point: even if we accept that the cotton gin is a technology "used for war", it most certainly was not the first nor the primary use of that technology (the first use was clothing).
        Veedrac 9y ago
        You're cherry-picking an irrelevant property of the argument. So what if, hypothetically, general AI is used nth rather than 1st in line for warfare? It's still AI used for warfare.
        9y ago
  - TeMPOraL 9y ago
    That's tautological. The very natures of technology and war mean that any and all technologies can, and will, be used for military, directly or indirectly. As you say, "If it can be used in society it can be used in war". But this is not a good argument against technology or research.

shusson 9y ago

As Rodney Brooks was recently quoted: "There are quite a few people out there who’ve said that AI is an existential threat: Stephen Hawking, astronomer Royal Martin Rees, who has written a book about it, and they share a common thread, in that: they don’t work in AI themselves" [1] [1] https://techcrunch.com/2017/07/19/this-famous-roboticist-doesnt-think-elon-musk-understands-ai/ https://techcrunch.com/2017/07/19/this-famous-roboticist-doe...

ericb 9y ago
Elon Musk works in AI--he's leading an attempt to make truly autonomous vehicles.
- shusson 9y ago
  Yeah it's a loose definition, but I think the point is Musk does not develop or research AI directly. He owns a company that is developing AI.
Aron 9y ago
That's gotta be some kind of pick and choose fallacy (availability bias?). Those two happen to be high-status individuals from old high-status science disciplines so they tend to get whatever they say amplified. I wouldn't use this kind of reasoning to figure out what opinion AI professionals as a class have, but I do know there are more celebrities outside of AI than inside of AI.
WheelsAtLarge 9y ago
It doesn't take much AI to create smart weapons. A smart drone can do a lot of damage. I don't even want to think about a level 5, completely automatic, car, which is what tech visionaries are aiming to create. Tech creators look at the bright side of their creations. They have tunnel vision and can't foresee the harm. If they can't others need to. The criticism that Musk does not know enough about AI ignores the fact that he is trying to create a fully automatic car with AI and has experience with its limitations. Yes AI(a catch all process for making computers do things that require intelligence when done by humans) is a process and it has no intelligence in itself but you can say the same thing about C++. It's just a language but look at what's possible to do with it.

hb3b 9y ago

Elon's tweet: https://twitter.com/elonmusk/status/889743782387761152 https://twitter.com/elonmusk/status/889743782387761152

dsacco 9y ago
I'm normally a huge fan of Elon Musk, but that tweet comes across as pretty hostile to me. Neither party seems to have critically backed up their beliefs, but I didn't get the sense that Zuckerberg was attempting to single out any individual for a lack of understanding.

ilaksh 9y ago

I don't think people like Musk or the others that he gets some of these ideas from, such as Kurzweil, are actually saying we shouldn't pursue advanced AI. Are they? I mean Musk helped fund OpenAI right? He is just saying that it is actually an existential threat or will be soon enough that we need to consider the problem seriously. I think the point of him mentioning that besides just worrying about it, is to motivate efforts like Neuralink that will allow for high-bandwidth communication between the brain and AIs. There is some misunderstanding on the part of Zuckerberg and others that saying that AI is an existential risk is the same thing as saying 'we should suppress AI research'. You would not think that was the implication if you understood from the perspective of Kurzweil/Musk/other people who have read Kurzweil's/Bostrom's book(s) or think the same. The idea is that we cannot possibly stop advanced human-like and then human-surpassing AI from developing. No matter how hard we try to slow down AI development. So when they say that AI can be dangerous, they do not mean that we should try to stop developing AI, since that is impossible. What they mean is that we should try to mitigate the risks to some degree. One way, which Musk and others are pursuing, is to try to create strong but friendly AI before some less friendly but equally potent variant becomes dominant. But if we do not develop some kind of high bandwidth interface with computers and AI that we control, eventually we should expect computer-based intelligence to significantly surpass our capabilities, to the point where we become irrelevant. If you disagree that there is any possibility of strong AI or any existential threat, what harm is there in pursuing strong but friendly AI, or in developing the BCIs (brain-computer interfaces)? The only 'harm' I can see is if this gets oversimplified as it has been to the idea of "AIs could be bad, we must stop AI" -- which they are not trying to say.

bredren 9y ago
Wasn't Elon's original point that he has current access to working AI demonstrations that personally frightened him, and served as the basis of his call for regulation?
- tim333 9y ago
  Kinda though his arguments seem largely from first principles that robots will get smarter than us so we'd better be careful what they get up to. Here's him chatting a bit at "National Governors Association Streamed live on Jul 15, 2017" https://www.youtube.com/watch?v=2C-A797y8dA&feature=youtu.be&t=48m9s https://www.youtube.com/watch?v=2C-A797y8dA&feature=youtu.be...

ExactoKnight 9y ago

No, it's Elon Musk whose understanding of AI is limited. I doubt Musk has ever actually built a neural network, or got close to coding one... because if he did, he would realize just how fragile and hard to generalize they are. Stacking a ton of data together to build predictive models that can produce intelligent-like outputs, is a far, far cry from a robot gaining consciousness or a sense of agency.

r_singh 9y ago
With Musk talking about AI the way he is, seems to me that he wants to use his reputation and the public's naivety to instill fear and have some control over the distribution of AI that works for general purposes...sounds like a great opportunity. If he stays consistent with what he says about AI, I think he might be successful in deterring people from using useful and reasonable AI for a while. I wish he would take the dangers of new features like Autopilot in a production car as seriously though.
- TeMPOraL 9y ago
  > If he stays consistent with what he says about AI, I think he might be successful in deterring people from using useful and reasonable AI for a while. What exactly did he say that would imply that? As far as I know (and I did watch all the recent interviews), his approach boils down to: - more open access to AI is less dangerous than it being developed in secret by corporations or governments; hence OpenAI - suggesting that regulators start paying attention (and explicitly advising them not to start regulating outright; just to form an agency that would "gain insight" into the field)
  - ExactoKnight 9y ago
    I do agree with this. One of the clearest inequalities emerging in A.I.-driven machine learning is how useless those techniques are unless you have access to a monstrously large amount of well categorized data. Currently you need to be one of the big four to have that access, and this is highly monopolistic.

cartyboi29 9y ago

My nigga Elon gonna pull up to zuckerfucks house with a 44

mncharity 9y ago

For a bit more thoughtful and nuanced discussion, the recent AI Now 2017 - Public Symposium[2] stream[1] (2.5hr) wasn't bad (at 1.5x speed). [1] https://www.youtube.com/watch?v=npL_UsK_npE&t=285 https://www.youtube.com/watch?v=npL_UsK_npE&t=285 [2] https://artificialintelligencenow.com/schedule/2017-symposium https://artificialintelligencenow.com/schedule/2017-symposiu...

richev 9y ago

Sounds like Musk has been reading the 1997 book written by my old Cybernetics professor Kevin Warwick (aka Captain Cyborg)... https://en.m.wikipedia.org/wiki/March_of_the_Machines https://en.m.wikipedia.org/wiki/March_of_the_Machines

bryanrasmussen 9y ago

That article just seemed to run out at the end, it seemed like there was some previous problem between the two related to the Shuttle? And then the article didn't describe it where I was reading.

nradov 9y ago

This debate seems so premature. It's hard to take seriously when so far no one has even come close to building an AGI equivalent to a mouse, let alone a human. Here is a broader set of perspectives from people working in the field. http://spectrum.ieee.org/computing/software/humanlevel-ai-is-right-around-the-corner-or-hundreds-of-years-away http://spectrum.ieee.org/computing/software/humanlevel-ai-is...

_archon_ 9y ago
If we have an AGI equivalent to a mouse, wouldn't it iterate itself to a human intelligence pretty rapidly? Safeguards and intelligent thinking now could be the difference between: Mouse -> cat -> shark -> Evil human mastermind -> problems and: Mouse -> cat -> dolphin -> helpful friendly human -> better world The end situation is highly theoretical, but a self-modifying AGI will think differently than a human does, so it's hard to predict what could happen with our existing preconditions. With such a broad range of potential divergence, and a steep difference in desirability of potential outcomes, it seems reasonable to ask some pointed questions sooner rather than later.
- nradov 9y ago
  No there's no evidence to indicate that would happen. Mice have been around for a long time. How much smarter is a mouse today relative to a mouse a million years ago? And why would a simple mouse level AGI have the intelligence to improve its own abilities?