back to article Warren Buffett voices AI fears, likens tech to atom bomb

You can add Berkshire Hathaway chief Warren Buffett to the list of folks worried about the implications of artificial intelligence on society. During the firm's annual shareholder meeting on Saturday, Buffett compared the technology's potential to the Manhattan Project and the arms race that followed. "I really don't know …

  1. may_i

    If I could post an image...

    but as I can't, you'll have to use your imagination.


  2. Dan 55 Silver badge

    "I really don't know anything about AI... Used in a pro-social way, it's got terrific benefits to society."

    I'm trying to think of a pro-social way of using "AI"... As far as I can tell it's just disruption² with all that entails.

  3. vtcodger Silver badge

    There's a storm coming?

    Thirty years ago most tech CEOs had some sort of grasp of technology. Back then, a committee of tech leaders asked about a new technology might actually have come up with something meaningful and maybe even useful. Today, the very clever geeks of the late 20th century have mostly been replaced by bean counters, marketeers (which is to say professional liars), and assorted scumbags.

    I have no more idea than Mr Buffett what the impact, if any, of AI will be. I fear that the best we can realistically hope for might be that AI fails miserably at pretty much everything non-trivial that it attempts. That folks lose faith in it, move on to some new fantasy, and AI is allowed to develop slowly and quietly into something with some actual utility and hopefully minimal catastrophic potential.

  4. Anonymous Coward
    Anonymous Coward

    Dystopian AI is no more inevetitable than the mass use of weaponized anthrax or the torment nexus

    Don't build it. Watch the people who could(it's a fairly short list), promptly punish anyone who tries well beyond an reasonable hope to gain.

    We don't have a global weaponized anthrax problem. That's partly because the usual rules of governments don't apply. Reasonably competent people jump to and start running the problem down. Those people get help from law enforcement pretty much anywhere, and even rogue states think really hard about if it's worth it. There have been a few incidents, but the enforcement regime is sufficient to keep the potential problem in check.

    So if we have to publicly flog every rich idiot that says if we don't build dystopian AI someone else will, so be it. Though to be fair in W.B.'s case a fine of 10% of his net worth might also get the point across. No need to be bloodthirsty after all.

    No deal for Zuckerberg. That plastic faced monster knows _exactly_ what he's trying to build.

    1. AI Hater

      Re: Dystopian AI

      And yet no one is doing it.

      We need to organize and save ourselves.

      So far there is only PauseAI

  5. Anonymous Coward
    Anonymous Coward

    Nut fudge

    Unchecked Albert Einstein Letter AInflation is the doom (and gloom) to watch-out for if I under-stood well ... (brought about by AI debt). Incarcerated economies are more immune to this than free ones ... but screw them (IMHO)!

  6. Bartholomew

    I do not far AI

    Fearing AI is stupid, it is already here and will not go away, fearing AGI on the other hand is not. AI can replace your job, AGI could result in human extinction.

    AI (Artificial Intelligence) = the use of learning and intelligence to take actions that maximize their chances of achieving defined goals.

    AGI (Artificial General Intelligence) = the ability to complete ANY task performable by a human on an at least equal level.

    AGI should be nearly possible very soon (current estimate is April 2024 on Deepsouth), but we are not there yet*. The electrical costs and storage to run a machine with enough SUPS (Synaptic Updates Per Second) will be totally insane.


    "In 2013, the (Japanese) K computer was used to simulate a neural network of 1.73 billion neurons with a total of 10.4 trillion synapses (1% of the human brain). The simulation ran for 40 minutes to simulate 1 s of brain activity at a normal activity level (4.4 on average). The simulation required 1 Petabyte of storage."

  7. amanfromMars 1 Silver badge

    Praise where praise is due, and the pure raw gems here deserve it in spades.

    "I really don't know anything about AI," the super-investor,Warren Buffett, admitted during the meeting. "Used in a pro-social way, it's got terrific benefits to society. But, I don't know how you make sure that happens any more than I know how to be sure that when you use two atomic bombs in World War II that you'd know you hadn't created something that could destroy the world later."

    Buffett also implied that just as with nuclear warfare capabilities, our hands are somewhat tied in the development of AI technologies in that if we don't adopt it, someone else will.

    Amen to all of that, Warren, with your concerns about the very real possibility of AI friends with benefits being equally capable of entertaining and enjoying radically hostile with deployment and employment of fundamental and almighty overwhelming existential threats to humanity's capitalising market players orders/disorders ....... a global systemic weakness and indefensible vulnerability for AI and 0day exploitation of what are essentially dinosaur operations enslaving future progressive freedom.

    And yes, AI technologies/methodologies/developments do have you captured and captivated ..... but one can buy into them in order to give them a reason for not extinguishing your co-opted participation in their Future Operating Systems, and present you a fabulous reason for continued living, adopted and adapting in a Brave New More Orderly World Order with Novel Otherworldly Ethereal Leadership[s]

  8. Splod

    He's not wrong!

    I agree AI is as big and as dangerous a development as nuclear power and like that it can go either way. But ... you cannot stop development so better make sure we stay in control.

    1. amanfromMars 1 Silver badge

      Re: He's not wrong!

      But ... you cannot stop [AI] development so better make sure we stay in control. ....... Splod

      :-) AI would ask you, Splod, if it thought for even a second there was any chance of that ever being possible, who be the we you have mentioned enabled with control of AI, and how do they command and exercise it to do as they would see fit and proper, rather than as AI realises is the way things are to be for a much brighter and greater future ‽ .

      And as you must surely hear loud and clear, there is a deathly silence from AI, wholly indicative of that fat chance being recognised and accepted as a forlorn hope and dismissed as impossible.

      1. N Tropez

        Re: He's not wrong!

        Interrobang! Upvote for amanfromMars1

  9. Anonymous Coward
    Anonymous Coward

    Corporate Weaseling

    “IBM CEO Arvind Krishna has been rather blunt about where this is headed at his business. Last spring, the exec boasted that as many as 30 percent of IBM's back office jobs could be automated with AI.

    Big Blue has since joined a consortium of tech giants that aim to identify which IT jobs are likely to be eliminated by the technology first, with the stated goal of retraining staff to fill jobs that AI isn't ready for”

    What complete IBM horseshit. The biggest ‘resource action’ yet is on it’s way:

  10. Conundrum1885


    Makes me wonder, would QC make AI look like a storm in a teacup??

    Being able to break hard encryption would be a game changer.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like