• 2 Posts
  • 2.4K Comments
Joined il y a 3 ans
cake
Cake day: 29 juin 2023

help-circle

  • I’ve noticed things getting worse personally. My main complaint is that people have started being much less charitable readers… A lot of people jump to conclusions that make me think “when the hell did I say any of that?”

    I’ve mostly just stopped replying at this point, no amount of clarification seems to help with these people, they’re just itching for a fight

    It’s definitely a minority, but it’s got me posting a lot less




  • The idea is that “AI” is used to detect possible targets

    Okay, but what does that mean? How does an llm do that exactly? What are the inputs and outputs?

    Do you actually know, or did propoganda get you? No shame in it, if you hear something enough and it can worm it’s way in without conscious evaluation

    I think neural networksthat highlight suspicious movements are fine, or at least as fine as weapons of war are fine in general. But ideally, that’s just battlefield information fed to humans, which isn’t how this is spoken about in the media

    But we’re specifically talking about grok here. How does grok fit into the process of firing a missile exactly, and what are the incentives built into the system for the people involved?

    You’ve got a decent grasp on the tech side, so I think if you consider the position of the stake holders you’ll come to the same conclusions.

    Put yourself in the position of the developers providing models, the brass and why they want this, and the MIC shareholders and what they’d lobby for. Think of these pressures, and ask yourself what kind of direction the project is pushed in

    There’s some pretty horrifying implications


  • What I don’t get is how a chatbot fires a missile…I mean you could give it a function call, but I don’t think we’re at that point of insanity just yet

    Maybe it could read through Intel and recommend targets, but there’s no way anyone is reviewing the research thoroughly, with the diligence appropriate before taking lives

    Can we stop the bullshitting and call the media out on this more? This is obviously just offloading responsibility to an entity beyond punishment

    A fourth grader could see the problem here… It’s not if AI is or isn’t competent, it’s that the people involved are incentivized to blindly trust the AI. That no one is taking responsibility for, so really it’s no one’s fault when schoolchildren are slaughtered