points by otabdeveloper4 22 hours ago

The LLM predicts next token one at a time. (Stochastically.) This is a literal truth.

Deal with it.

layla5alive 17 hours ago

It's a literal truth that predicting the next token one at a time does not preclude intelligence on the other side of the decode function. Deal with it.

  • otabdeveloper4 2 hours ago

    Yes, for a certain vague notion of "intelligence". Even homes, televisions and watches are "smart" now.