We have certainly seen examples of failure in experts during the COVID-19 pandemic. The first, and perhaps most famous example, is the death estimates from the Imperial College model. One of the fundamental principles of computational model is that the model is only as good as the inputs put into it and the assumptions made; in most computational models, few have the knowledge and skills required to use them effectively and understand their limitations. The answers helped it make the news, not the accuracy or level of uncertainly. This led initial experts and politicians to make very wrong assumptions; for instance, anticipating extreme hospital space shortages led to recovering COVID-19 patients being sent into nursing homes while still contagious. I would also argue testing and the lateness of the vaccine (it easily could have been administered much earlier absent the aggressive regulatory environment) was to a large degree due to the recommendations of experts; they argued their test was better (it was probably worse), reprimanded those who used it, and said challenge trials (that could have proven the vaccine sooner) were inherently unethical.