Am I the only one slightly annoyed by the glib usage of "exponential" in every AI discussion?
I'd like to know what exactly is growing exponentially. Compute used for training certainly is. I'm not sure that capabilities are because we don't have a great way of actually benchmarking these models any longer.