AI providers really need to set expectations correctly. They are getting into trouble by allowing people to think the answers will be correct. They should be

AI providers really need to set expectations correctly. They are getting into trouble by allowing people to think the answers will be correct. They should be stating up front that AI tries to be correct but isn't always and you should verify the results.