The Pentagon has its eye on the leading AI company, which this week softened its ban on military use.

you are viewing a single comment's thread
view the rest of the comments
[–] 113 points 2 years ago (3 children)

I can't wait until we find out AI trained on military secrets is leaking military secrets.

  • source
  • hideshow 6 child comments
  • [–] 24 points 2 years ago (1 child)

    I can't wait until people find out that you don't even need to train it on secrets, for it to "leak" secrets.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 6 points 2 years ago (1 child)
  • [–] 7 points 2 years ago (1 child)

    Language learning models are all about identifying patterns in how humans use words and copying them. Thing is that's also how people tend to do things a lot of the time. If you give the LLM enough tertiary data it may be capable of 'accidentally' (read: randomly) outputting things you don't want people to see.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 14 points 2 years ago (1 child)

    I mean even with chatgpt enterprise you prevent that.

    It's only the consumer versions that train on your data and submissions.

    Otherwise no legal team in the world would consider chatgpt or copilot.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 4 points 2 years ago

    I will say that they still store and use your data some way. They just haven't been caught yet.

    Anything you have to send over the internet to a server you do not control, will probably not work for a infosec minded legal team.

  • source
  • parent