AI Companies Acknowledge that They Might Kill You in the Near Future

My newest article takes a look at what some people at the leading AI companies have said about the riskiness of the technologies they’re trying to build. Noted figures from Sam Altman and Ilya Sutskever (OpenAI) to Dario Amodei (Anthropic) and Shane Legg (DeepMind) have acknowledged that it may be impossible to control artificial general intelligence, or AGI, and that the likely — or at least possible — outcome of this is the total annihilation of humanity.

You do not need to believe that AGI could actually kill everyone on the planet to be profoundly alarmed by this. People running projects backed by billions and billions of dollars are saying that their tech might kill you, your family, and all your friends in the near future — perhaps in the next few years. It is difficult to convey without cursing how completely unacceptable that is. If I were to tell you, the reader, that I’m up to something that might murder you in the near future, then even if you think I might be delusional or lying, you should still be rather alarmed and disturbed by such a remark. But it’s not me saying this to you — it’s immensely powerful people within extremely powerful industries saying that. That is. Completely. Unacceptable.

Read more here:

https://www.truthdig.com/articles/the-madness-of-the-race-to-build-artificial-general-intelligence/