you are viewing a single comment's thread
view the rest of the comments
[+] -23 points 14 hours ago* (last edited 6 hours ago) (4 children)

AI labs bad, grr, but...

[strikethrough]This is a clear misrepresentation on the part of the twitter poster, and propagated by OP.

This is obviously referring to AI / the AI industry as a whole.[/strikethrough]

edit: I missed some context in my reading, and I hit send without my usual double-check when family hell broke loose.. They ARE focused on their own potential liability issues here, but it is of course a common problem that all AI service providers face (and potentially even those releasing open-weights models). Ultimately, they want/need friendly legislation across multiple countries that lets them manage or eliminate legal risk.. and everyone involved would appreciate more legal certainty. It reminds me of the Sun Java license that said you cannot use it for anything nuclear related.

  • source
  • hideshow 4 child comments
  • [–] 9 points 12 hours ago

    You are correct, it's also true that the cost of an AI industry collapse would be beyond the insurance industry to indemnify. Which honestly isn't a problem. Some things need to be allowed to fail.

  • source
  • parent
  • [–] 3 points 10 hours ago* (last edited 10 hours ago)

    You idiot. They were referring to Anthropic's models explicitly, but it could be a downstream issue (it is also true of every company providing AI models). If the law were to ever hold them accountable, there would potentially be huge liability which would be uninsurable.

  • source
  • parent