Beskrivning av hur AI (LLM) skapas.
Citat från nobelföreläsning 2024
Geoffrey Hinton, nobelpristagare i fysik (ännu inte i Physiology and Medicine)
"In the forward path, you are communicating activities in neurons to later layer neurons.
In the backward path, you are conveying sensitivities, a different quantity altogether,
that makes backward propagation rather implausible as a theory of how the brain works.
So when Terry came up with this theory, this learning theory for Boltzmann machines,
we were completely convinced that must be the way the brain works.
We decided we were gonna get the Nobel Prize in Physiology and Medicine.
It never occurred to us that, if it wasn’t how the brain works (ironiskt skämt),
we could get the Nobel Prize in Physics. (skratta här)"