> I first aired my concern in March 2022, in an article called “Deep learning is hitting a wall.”

In March 2022, GPT-3 was state of the art. Why should anyone care what he's saying now?

because i correct foresaw quite a lot. you should actually read the paper.