Thread

1. ck2 · 1791063240 · focus · HN ↗

   I've found the best way to explain what LLM "AI" actually is to people who know nothing about computers despite hunching over their smartphone 12 hours a day

   Is to point how typing corrections and suggestions work because they get that and how it makes mistakes "damn you autocorrect"

   Then explain that's almost how words are stitched together in "AI"

   And how pixels are manipulated to make photos and video

   Just on an exponential level over autocorrect with every existing text document, scientific study, every photo and video fed in instead of words to autocorrect

   And why they make mistakes and hallucinate if they are not supervised and harnessed

   We are a long way from AGI, at least not this decade

   (and someone really needs to explain that to the orange idiot too)

   1. b0rtb0rt · 1791137728 · focus · HN ↗

      how does your brain decide what words to say and what order to say them in? how do you know that it’s not the exact same thing you’re describing but using the context of your own memories?

Unofficial Hacker News client; not affiliated with Y Combinator.