> LLM will perform poorly

We've been seeing examples of this constantly since 2022. How many more do we need?