Ah, so you think there’s a point where actually bloat slows because we eventually can’t keep up with demand for compute? I guess this might be happening with L

Ah, so you think there’s a point where actually bloat slows because we eventually can’t keep up with demand for compute? I guess this might be happening with LLMs already