I don't think people are claiming he was not experimenting as much as they are claiming he was overtly optimistic about the outcome. It seemed like he went in with the notion that AIs are somehow thinking machines. I don't think that's an objective sentiment. An unbiased researcher would go in without any expectation.