These kinds of psychology experiments are my pet peeve and not just the ones about AI.
Expose a group of people to something for 15 minutes, then measure something artificial in the next 15 minutes, and then use these results to make a headline about the long-term effects of that something.
