How might language models be relevant to AI alignment?1 min read<answer goes here> Will the first AGI be an LLM?