I was just answering this question. LLM logic in weights is fundamentally from machine learning, so yes. Wasn't really saying anything about the article.