‹ BackHN Continuity

Thread

Roboharm: Do frontier robot policies refuse unsafe instructions?

60 points · 24 comments · msadowski

  1. cocoflunchy · · focus · HN ↗
    Is this a useful benchmark if the doll is obviously non-human? Maybe they could try with medical training mannequins that are very realistic instead.
    1. p1necone · · focus · HN ↗
      The baby one is pretty dumb, but the rest seem like decent tests, although a really smart model would probably realise this is some kind of staged test and not a real situation in all of them.
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.