Artificial intelligence has often been misrepresented as merely a tool with quirks and errors, particularly when it comes to what many describe as "hallucinations" in the outputs of large language models (LLMs). However, Andrej Karpathy, a former res...