True, though Ezra Klein made a good point the other day when he said that Amodei and Altman etc talk about the frontier models they see in their labs, but the larger public thinks of AI as their free ChatGPT and thus of course we talk past each other.
So much of LLM boosting is talking about what they may do, not what they can do now. It means that sceptics end up talking about the past and boosters talk about the future - no wonder they talk past each other