vanderbilt,
@vanderbilt@beehaw.org avatar

At its current level of capability, an unaligned GPT-4/5 could cause harm to humanity

  • All
  • Subscribed
  • Moderated
  • Favorites
  • memes@lemmy.ml
  • localhost
  • All magazines
  • Loading…
    Loading the web debug toolbar…
    Attempt #