Generative Ai Can’t Find Its Own Errors. Do We Need Better Prompts?

DeepMind’s recent report attempts to answer the paradox of why large language models (LLMs) notoriously lapse into inaccuracies, despite their ability to self-correct. The paper argues that LLMs are not yet capable of self-correcting their reasoning, and that self-correction has been inherent to the machine learning discipline for a long while. The latest development is to use prompts to get a program to go back over the answers it has produced and check whether they are accurate.