Continual Learning has already been solved, it just requires the model weights to be running locally on your own hardware so big labs are avoiding the topic altogether Local / Opensource AI will win Inevitable

→ View original post on X — @theahmadosman