Anthropic is one of the early adopters of chat GPT. The company says it's not going to release a new level of AI capability until other companies have already reached that threshold. Anthropic has published really interesting research about understanding these giant models and they are dedicated to finding new methods for AI safety.