Since AI remains in the early stages of testing, it is important to have external monitoring of model results. Large language models should be tested through third parties and results made available to the public. That will help people understand how AI is performing and which applications may be especially problematic. (2023) source Unverified