Australian government trial finds AI is much worse than humans at summarizing
arstechnica.com
Llama2-70B failed to capture "complex context," but updated models might do better.
  • @[email protected]
    4•4 months ago

    Yeah! You ever try to sort through the garbage the AI “thinks” is relevant information? It’s awful and then you have to retype and ask it again and then it still might be imagined and then you go on a hun for what it’s source was cause it doesn’t tell you and
    By that point you could have just read the original paper and not looked like an idiot quoting a barely functional autocomplete program.

    • Annoyed_🦀
      2•4 months ago

      But hey, it create the “prompt engineering” and “AI proofread” job, all good!