• [object Object]@lemmy.ca
    link
    fedilink
    arrow-up
    14
    arrow-down
    1
    ·
    edit-2
    18 days ago

    These things are the worst.

    • the agent has zero memory
    • the agent has zero autonomy
    • the agent will not disclose to you what is tools are, and you cannot tell when it is hallucinating or exaggerating
    • you now are sharing your voice with AI companies with zero consent
    • you have no way to verify they have actually done something or scheduled a follow up
    • this is AI running in a cost centre. Its job is to NOT connect you with a human. Therefore any failure actually pads its success metrics.

    Fuck this.

  • reddig33
    link
    fedilink
    English
    arrow-up
    4
    arrow-down
    1
    ·
    18 days ago

    For me, it’s going to depend on how this is being utilized. If it simply asks you for a summary of your problem, creates a case and transfers you to the appropriate human being trained for your product, it could be helpful.

    If it’s a glorified chatbot preventing you from talking to anyone, then fuck that.

  • paraphrand
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    1
    ·
    18 days ago

    Not cool 😔. I expect it’s just for routing, tho.