“John Schulman and his team at OpenAI fully pivoted their research focus to language models only after the release of GPT-3, as GPT-2's capabilities had not yet convinced them of the technology's revolutionary potential.”