Ask HN: What are some collective blind-spots amongst people working on AI?

i.e. Many of the people share similar backgrounds, so would expect them to have common blind-spots.