How is this a counter argument that LLMs are marketed as having intelligence when it’s more accurate to think of them as predictive models? The fact that humans are also flawed isn’t super relevant to a $200/month LLM purchasing decision.