Opinions make the post-truth world go round! Opinions are influential. Therefore, it is essential to ensure that opinion polling is conducted properly and avoids common pitfalls such as biased question formats. What if we could get representative polls of public opinion in a fast and automated way? Would it be possible to remove humans from such a process? This research tackles the first stage of such a project by building a pipeline to automatically generate opinion questions. We evaluate the use of GPT-3, a large language model, in generating opinion questions on trending topics. Ultimately, we answer the research question: how accurately can human evaluators distinguish between human and GPT-3 generated opinion questions?