Flaws in the LLM Automation Narrative

George Perrett, Javae Elliott, Jennifer Hill +1

Large Language Models (LLMs) are increasingly described as performing at the level of human experts on knowledge economy tasks. These claims are primarily based on how LLMs perform…