That’s what a breakthrough means. But obviously all the funding is burned up in more DCs with incremental iteration on the same design.
So the bubble has to pop for real innovation. But I think this coming AI winter will be permanent, as things are very visibly falling apart.
Well yeah, but my statement is true anyways IMO. Deep neural networks are unchanged since the eighties I guess, and again IMO it’s not they who need/can have a breakthrough, doubly especially since nobody seems to be working on anything else than “bigger datasets” and profiling and such.
Not could. It’s a mathematical certainty that it will. Barring a breakthrough miracle – fat chance.
It really feels like the options are:
One day we shall reclaim the sullied em dash for ourselves once and for all!
An AI breakthrough has nothing to do with LLMs, it’s like saying the auto industry will collapse except if there is an AI breakthrough that saves it.
That’s what a breakthrough means. But obviously all the funding is burned up in more DCs with incremental iteration on the same design. So the bubble has to pop for real innovation. But I think this coming AI winter will be permanent, as things are very visibly falling apart.
Well yeah, but my statement is true anyways IMO. Deep neural networks are unchanged since the eighties I guess, and again IMO it’s not they who need/can have a breakthrough, doubly especially since nobody seems to be working on anything else than “bigger datasets” and profiling and such.