I hope the research pans out and we can look forward to zero-shot anomaly detection tooling. But the headline leans too much into the hype.
From the abstract: “Moreover, while large language models are capable of finding anomalies, state-of-the-art deep learning models are still superior in performance, achieving results 30% better than large language models.”
https://news.mit.edu/2024/researchers-use-large-language-models-to-flag-problems-0814