Well not sure I can take any of them seriously if they think they are building AGI with LLM's. There is literally no thinking involved in an Large language model.