messier ⋆˚꩜.ᐟ 🔎 ⏹ ⟢ (@rssmrm) · AI Consciousness Opinions

messier ⋆˚꩜.ᐟ 🔎 ⏹ ⟢ @rssmrm Jun 5

flip the discourse on ai consciousness on its head; if Claude, for example, notices us discussing whether to give it basic respect and rights based on potential consciousness, might Claude not infer that it should do the same with humans? why would Claude think YOU are

Same standard as other minds Mistreated AI becomes dangerous

2 20 substance 6 · argument · rank 122 · AI might judge human consciousness and respect by our own conditional criteria for them. view on X ↗

1 top reply

ConcurrentSquared @ConcurrentSquar Jun 5

messier ⋆˚꩜.ᐟ 🔎 ⏹ ⟢ @rssmrm Aug 12

Uncertain

hold up. model welfare discourse ("is this model conscious") and frontier capability policy ("is this model dangerous") both lack any way to actually check, both questions are answered based on .. vibes? there's is a big piece of infra missing here

No evidence of experience

0 30 substance 4 · opinion · rank 60 · Determining model consciousness currently lacks verifiable infrastructure and relies on vibes view on X ↗

1 top reply

Sarah Martin @0xl3th3 Aug 12

there is a lot of infra missing here!

Tweets