My problem with the test is that I’m not sure it’s really testing what it purports to be testing. It says that it’s testing your ability to discern misinformation from looking at titles, but I think what it’s really testing is your ability to differentiate human written titles and AI generated titles. AI generated titles could be truthful and human written titles could be utter bullshit and without checking the credibility of the source or reading the article, you’re not necessarily going to know which is which unless you already know something about the topic.

That said this is an interesting experiment that I predict will not have the same results when LLMs become more advanced.