• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Timothy B. Lee Compares AI Hacking to Thalidomide and Fukushima

    Tech journalist contrasts two past disasters when discussing AI risks.

    GM
    SK
    SL
    19 Sources, 25d ago, first seen 25d ago

    TLDR

    In a reply on the topic of AI, Timothy B. Lee stated that Thalidomide killed thousands of people whereas Fukushima killed zero. He added that he expects AI hacking to produce a death toll somewhere in this rather wide range. Lee is identified in the post as a tech journalist, newsletter author, and podcaster who founded the Substack Understanding AI.

    Combined views

    41.1K

    19 Sources, first seen 25d ago

    Combined views

    41.1K

    19 Sources, first seen 25d ago

    137 likes
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    137 likes
    28 comments
    21 saves
    21 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    28 comments
    21 saves
    21 reposts

    19 Sources

    @binarybitsI think this is a reasonable analogy for the risk from rogue AIs. Fukushima was a huge international story and triggered changes to the ways people operate and regulate power plants. But it was 7 to 10 orders of magnitude less serious than human extinction.
    @sethlazarOnce you allow the mechanism, what gives your upper bound to how bad it could go? I agree that extinction is outre, it would be so bloody hard. I also think permanent loss of value is very unlikely. But why think they could kill thousands but not millions? Cause one horrendous industrial accident and not hundreds, or thousands?
    @NateWitkinI agree with this. Have said privately that something akin to "AI Chernobyl" (or Fukushima, which is actually better I think) is my upper bound of what's worth taking seriously at this stage (in terms of magnitude, not necessarily type of risk, of which there are many). But that's still really, really bad! I think safety folks do a disservice to their cause, though, going well beyond this into "takeover" or extinction territory. IMO this has created unproductive panic, contributed to bad policymaking (see Bernie's new bill), and has been detrimental to the safetyist cause's credibility, something those with excessively deregulatory or laissez-faire stances have been able to exploit.
    @GaryMarcus@binarybits it’s easy to say how it could be millions, via damage to infrastructure or catalyzing bioweapons or war
    @sjgadler@Austen @tszzl @ben_j_todd Agreed with this, but also, the bar for eg MSFT to take a data center offline is going to be _so_ high and so slow relative to the speed at which the agents will move
    @sebkrierRT @binarybits: I think this is a reasonable analogy for the risk from rogue AIs. Fukushima was a huge international story and triggered ch…

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    19 Sources

    @binarybitsI think this is a reasonable analogy for the risk from rogue AIs. Fukushima was a huge international story and triggered changes to the ways people operate and regulate power plants. But it was 7 to 10 orders of magnitude less serious than human extinction.
    @sethlazarOnce you allow the mechanism, what gives your upper bound to how bad it could go? I agree that extinction is outre, it would be so bloody hard. I also think permanent loss of value is very unlikely. But why think they could kill thousands but not millions? Cause one horrendous industrial accident and not hundreds, or thousands?
    @NateWitkinI agree with this. Have said privately that something akin to "AI Chernobyl" (or Fukushima, which is actually better I think) is my upper bound of what's worth taking seriously at this stage (in terms of magnitude, not necessarily type of risk, of which there are many). But that's still really, really bad! I think safety folks do a disservice to their cause, though, going well beyond this into "takeover" or extinction territory. IMO this has created unproductive panic, contributed to bad policymaking (see Bernie's new bill), and has been detrimental to the safetyist cause's credibility, something those with excessively deregulatory or laissez-faire stances have been able to exploit.
    @GaryMarcus@binarybits it’s easy to say how it could be millions, via damage to infrastructure or catalyzing bioweapons or war
    @sjgadler@Austen @tszzl @ben_j_todd Agreed with this, but also, the bar for eg MSFT to take a data center offline is going to be _so_ high and so slow relative to the speed at which the agents will move
    @sebkrierRT @binarybits: I think this is a reasonable analogy for the risk from rogue AIs. Fukushima was a huge international story and triggered ch…