A critique of overhyped AI agent implementations, pointing out that some supposedly innovative LLM agent systems are simply while loops running unit tests. The author dismisses the novelty of such approaches, comparing them to basic programming constructs that have existed for years.
•1m watch time
Sort: