Slashdot Mirror


Elon Musk: Humans Need To Merge With Machines Else They Will Become Irrelevant in AI Age (cnbc.com)

Billionaire Elon Musk is known for his futuristic ideas. So it didn't come as a surprise when on Monday at the World Government Summit in Dubai, he predicted that over time we will see a "closer merger of biological intelligence and digital intelligence." He added, via a CNBC report: "It's mostly about the bandwidth, the speed of the connection between your brain and the digital version of yourself, particularly output." Musk explained what he meant by saying that computers can communicate at "a trillion bits per second", while humans, whose main communication method is typing with their fingers via a mobile device, can do about 10 bits per second. In an age when AI threatens to become widespread, humans would be useless, so there's a need to merge with machines, according to Musk. "Some high bandwidth interface to the brain will be something that helps achieve a symbiosis between human and machine intelligence and maybe solves the control problem and the usefulness problem," Musk explained.

7 of 251 comments (clear)

  1. Solve for Greed first. by geekmux · · Score: 5, Interesting

    AI can and probably will ultimately make the concept of human employment extinct. And we probably need to accept that fact faster than any prediction, given the speed at which technology has accelerated just in the last couple of decades.

    Humans need to first solve the problem of Greed. Otherwise, the chasm that separates the AI owners and wealthy overlords from the rest of the human race will continue to grow, and they won't give a shit about the demise of the irrelevant meatsack masses.

  2. He does have a point... by mspohr · · Score: 4, Interesting

    Most of the comments so far have been personal attacks on Musk. I guess this is par for the low level of discourse here.
    However, I'd like to see some discussion of his statement.
    Would a better connection between humans and machines be beneficial?
    What would be the benefits/ problems?
    How could this be achieved?
    Come on, folks, I have seen much better from you in the past.

    --
    I don't read your sig. Why are you reading mine?
    1. Re:He does have a point... by thinkwaitfast · · Score: 3, Interesting

      Great minds discuss ideas; average minds discuss events; small minds discuss people.

    2. Re:He does have a point... by AthanasiusKircher · · Score: 3, Interesting

      However, I'd like to see some discussion of his statement.
      Would a better connection between humans and machines be beneficial?
      What would be the benefits/ problems?
      How could this be achieved?

      To discuss something meaningfully, you need to have a freakin' clue how it might work. At this point in time, we don't. We don't know how the brain works. We don't have anything close to strong AI. The best interfaces we're looking at now are stuff like moving artificial limbs or whatever. To speculate on what might happen IF we could all of this would be sort of like walking up to Isaac Newton in the 18th century and saying, "Sir Isaac, what problems do you think will occur with the internet next year? What will the major benefits/problems be of new advances?" Even if you explained the basic idea of the internet to Newton, I doubt he'd have enough perspective to meaningfully debate what might happen.

      But, having put forth that disclaimer, I'll just note a few complete speculations in response to Musk. First is that his argument seems premised on the idea that a faster interface from brain to world would be beneficial to humans. Maybe it would. OR maybe our brains are somewhat limited in maximum input/output in ways that we can't really understand yet because we've never tried what he's proposing. Typing is about the right "speed of thought" for me to create coherent text. I've tried dictating, and I need to pause, correct, and reword too much for it to be useful to me. That seems to be how my brain works... although if I really needed to, I probably could retrain myself to dictate better.

      But what if you increased my potential output by 10-fold, 100-fold, a million-fold. Would that actually be useful for me to interact with the world better or faster? Or would it just result in gibberish because my brain literally can't adapt to working much faster than it already does in USEFUL output? Or maybe the plasticity of the adult brain isn't enough to adapt -- so we try hooking up infants from birth with these things. Maybe it works... or maybe it just drives them to be insane or to have other brain development that effectively renders them LESS functional than "normal" humans. Not saying this WILL happen, but it's a possibility when you're talking about an interface with absolutely NO IDEA on what specs might work. Human brains have spent millions of years evolving into what they are to work efficiently at the speeds they do. Just because you could theoretically hook up a device to increase input/output doesn't mean the brain can actually change and adapt enough to make use of the throughput meaningfully.

      Also, I think it's important to note in a discussion like this that one of the PRIMARY hallmarks of human intelligence is FORGETTING. One of the things that makes humans so much better than machines is our ability for abstraction -- finding larger patterns so we don't have to parse the "stream of consciousness" directly all the time. And then we sleep, and our brains revisit the memories of things that we've evolved to assimilate as "important" data, while we forget millions of random little details of our day at the same time.

      Effectively, we take a very TINY percentage of the "noise" that is input into our brains and actually remember it in any detail, mostly through complex pattern-matching that we're only even beginning to emulate in specific cases with computer algorithms. But the point is that there's only so much that we CAN assimilate into our brains -- and that goes not just for memories, but for new skills or whatever. (Think about when you've tried to learn a skill by "cramming" for a full day or two vs. when you've done practice for a few minutes/day over weeks or months. Your brain needs the "downtime" to assimilate new skills... increasing input or output seems unlikely to make that process faster.)

      My speculation is that Musk's idea is rather pointless for somehow keeping humans "relevant" or whatever. IF w

  3. ...and the benefits would be...what exactly? by holophrastic · · Score: 4, Interesting

    alternatively: humans only need to communicate at 10 bits, we don't need a trillion bits per second to enjoy life.

    But really, isn't the trick to do less, not more? I ain't no worker-bee. I'm jealous of my pet dog sitting on the couch all day while I work at a desk. I want his life -- it's called retirement.

    Productivity is the goal of business. Laziness is the goal of life. I've worked hard to be this lazy.

  4. Re:Elon Musk: What's this guy smoking? by The+Real+Dr+John · · Score: 3, Interesting

    Yeah, rich people don't get a lot of reality checks from their hired help. I just wish people were better at differentiating between Hollywood and real life. I hope Elon has a computer surgically implanted in his brain to show everyone how dumb an idea it is.

    --
    A brain is a terrible thing to waste... Mind? That's debatable.
  5. Re:language by flink · · Score: 3, Interesting

    I can't find the source, but I remember reading that speakers of lower entropy languages simply end up speaking faster than speakers of higher entropy languages. E.g. English on average is spoken more slowly than Spanish by native speakers because English has less redundancy, so errors are more likely to affect received meaning. Overall spoken information rate (bps) remains pretty constant.