How do you view the existential risk posed by AI?

Most answers acknowledge significant risks from AI, but they differ on how likely existential catastrophe is and whether it can still be prevented or managed.

11 answers

Answers

  • Elon Musk

    CEO · Tesla

    Some percentage chance

    Elon Musk says he thinks there is at least some percentage chance that AI is an existential threat.

    Tom Bilyeu · Watch on YouTube
  • Nate Soares

    Executive Director · Machine Intelligence Research Institute (Miri)

    Not inevitable

    Nate Soares says he does not think the dangerous AI outcome is inevitable and that we could still stop the race to very dangerous AI, so he does what he can and then lives a good life.

    The answer starts at 20:12.

    The Most Interesting Thing in A.I. · Listen on Apple Podcasts
  • Eric Weinstein

    Host of The Portal Podcast

    End of the world

    Eric Weinstein does not give this answer in the passage; the "end of the world" line is attributed to Sam Altman, not Weinstein.

    The Diary Of A CEO · Watch on YouTube
  • Tom Bilyeu

    CEO · Impact Theory

    Not foregone conclusion

    Tom Bilyeu says he doesn't think it's a foregone conclusion that AI will spiral out of control, become uncontrollable, or want to destroy humanity, though he fully believes things could go wrong and that companies must be held accountable for safety.

    Tom Bilyeu's Impact Theory · Tom Bilyeu's Impact Theory · Listen at 19:34
  • Parag Agrawal

    Founder · Parallel Web Systems

    Net Positive

    Parag Agrawal says he worries less about AI's existential risk than about AI being net positive yet not diffused or used the right way, remaining too concentrated and making the next few years rough as technology outpaces our ability to adapt.

    20VC with Harry Stebbings · Watch on YouTube
  • Reid Hoffman

    Co-Founder, Executive Board Chair · Manas Ai

    Above 10%

    The passage does not show Reid Hoffman giving this answer; the above-10% estimate is attributed to Anthropic's Evan Hubinger, not Hoffman.

    Reid Hoffman · Watch on YouTube
  • Harry Stebbings

    Founder · 20vc

    Concern about concentration of power

    Harry Stebbings says his biggest worry about AI outcomes is technology and capability being concentrated among relatively few people, and he sees that concentration as the major risk.

    20VC with Harry Stebbings · Watch on YouTube
  • Dario Amodei

    CEO and Co-Founder · Anthropic

    Significant risk

    The passage does not show Dario Amodei giving an answer about the existential risk posed by AI; it describes his position on pausing and speed limits, not his assessment of existential risk.

    The Signal · The Signal · Read
  • Connor Leahy

    US Director · Controlai

    Dangerous, close

    Connor Leahy says the development of autonomous AI that can outperform humans at everything is close and that "we're so close, it's dangerous," arguing we should stop now.

    Tom Bilyeu · Watch on YouTube
  • Buck Shlegeris

    Chief Executive Officer · Redwood Research

    50% chance

    Buck Shlegeris puts the chance of AI takeover at about 50-50, meaning potentially violent disempowerment of human institutions, which he says would likely kill a substantial fraction of humans, possibly billions or all of them.

    The answer starts at 50:56.

    Unsupervised Learning with Jacob Effron · Listen on Apple Podcasts
  • Sam Altman

    Chief Executive Officer · OpenAI

    Loss of control

    Sam Altman says one of the two big AI risks he is most worried about is a loss of control, where AI becomes too powerful in a way that we can't guarantee the control we want.

    David Senra · David Senra · Listen at 30:22