Luís Silva on Nostr: nprofile1q…6gr2u Just a disclaimer: I didn't RTFA - but it seems pretty impossible ...

nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqpc27tkltk9hl84ru7528x05xge763dallklnf6zkpg8walmgcgqs26gr2u (nprofile…gr2u) Just a disclaimer: I didn't RTFA - but it seems pretty impossible without any feedback loop, aka, introspection, with self learning capabilities. As they are currently, they can just simulate the thought process within a limited context. It's a great simulation though, and a significant leap into AGI,. but we aren't there yet, I don't even know if this is the right way to achieve it. One thing is certain, the context size and the "inference" surpasses a big chunk of human population, seriously, all of this by just predicting the next token...