15 pointsby ok12345620 hours ago2 comments
  • karmakaze13 hours ago
    > The default model is Claude Haiku 3.5, but users can also choose to use Claude Sonnet 3.5 v2 and Meta LLaMa 3.2, depending on the task.

    > The memo also includes a warning: “Do not type or paste federal nonpublic information (such as work products, emails, photos, videos, audio, and conversations that are meant to be pre-decisional or internal to GSA) as well as personally identifiable information as inputs.” Another memo instructs people not to enter controlled unclassified information.

    Wat! All that stands in the way of leaking private personal information is a memo to employees.

    I would have thought that it would use a government-private instance of xAI or something. Maybe that's the play, when info leaks then xAI can be brought in to save the day when the accident waiting to happen does.

    • verzali5 hours ago
      That's pretty standard unfortunately. My company also issued us a warning not to enter classified or confidential data into LLMs, but there's nothing (other than the threat of being caught, fired, fined, or imprisoned) actually stopping me from doing it.
    • out-of-ideas8 hours ago
      ha, sadly, this is not shocking to me. there are a lot of companies with this exact same memo and deal with pretty sensitive information (and that is all the details im giving out).
  • manbart19 hours ago
    Another grift from the Musk admin. The bullshit is piling up so high you need wings to stay above it