I have many conversations with people about Large Language Models like ChatGPT and Copilot. The idea that “it makes convincing sentences, but it doesn’t know what it’s talking about” is a difficult concept to convey or wrap your head around. Because the sentences are so convincing.

Any good examples on how to explain this in simple terms?

Edit:some good answers already! I find especially that the emotional barrier is difficult to break. If an AI says something malicious, our brain immediatly jumps to “it has intent”. How can we explain this away?

  • DaDragon@kbin.social
    link
    fedilink
    arrow-up
    6
    arrow-down
    1
    ·
    2 months ago

    The short hand answer I’d try to give people is ‘it’s statistics’. Based on training data, there’s a certain chance of certain words being in proximity of each other. There’s no reasoning behind placement, other than whatever pattern is discernible from known situation.