https://futurism.com/the-byte/government-ai-worse-summarizing

The upshot: these AI summaries were so bad that the assessors agreed that using them could require more work down the line, because of the amount of fact-checking they require. If that's the case, then the purported upsides of using the technology — cost-cutting and time-saving — are seriously called into question.

  • impartial_fanboy [he/him]
    ·
    14 days ago

    Maybe stop ignoring entire fields of research that, to this date, are still figuring out what biological brains are doing and how they are doing them instead of just nodding along to what you already want to believe from people that have blinders for anything outside of their field (computers, in this case).

    Well first, brains aren't the only kind of intelligent biological system but they aren't actually trying to 1 for 1 recreate the human brain, or any other brain for that matter, that's just marketing. The generative side of LLM's is what gets the focus in the media but it's really not the most scientifically interesting or what will actually change that much all things considered.

    These systems are absolutely fantastic at finding real patterns in chaotic systems. That's where the potential lies.

    It's like if people were trying to develop rocketry to achieve space travel, but you and yours were smugly stating that this particularly sharp knife will cut the heavens open, just you wait.

    More like trying to go to the moon with a Civil War era rocket, it is early days yet. But progress is insanely quick.

    • UlyssesT [he/him]
      ·
      14 days ago

      These systems are absolutely fantastic at finding real patterns in chaotic systems. That's where the potential lies.

      No arguments there; my issue is the marketing bullshit that wants to call them 1:1 "artificial intelligence" which is an insult and a dismissal of actual ongoing artificial intelligence research projects.

      More like trying to go to the moon with a Civil War era rocket, it is early days yet. But progress is insanely quick.

      My metaphor was heavy handed, I know. Maybe I should have said it's like trying to fire a bullet at the moon and just expecting more and more gunpowder to do the trick instead of considering a different approach using chemical propulsion.