Inference for the same results has been dropping 10x year over year[0]
Sure, but "the same results" will rapidly become unacceptable results if much better results are available.
loading story #47680280
loading story #47680323
loading story #47680222