futurism.comOpenAI Research Finds That Even Its Best Models Give Wrong Answers a Wild Proportion of the TimeOpenAI has released a new benchmark dubbed "SimpleQA" to measure the accuracy of its AI models. The results are damning.