“Dario Amadei developed the concept of "scaling laws" while at OpenAI, which predicts that large language models improve by increasing data and computing power, even if the underlying algorithm remains unchanged.”