That is what I kept thinking while reading the article too. 1) The YouTube link from the article mentions that most children let the robot pass, while the arti

That is what I kept thinking while reading the article too. 1) The YouTube link from the article mentions that most children let the robot pass, while the article reads as "100% of children are evil for no purpose". 2) Another comment mentioned the survey results possibly being skewed, because the children give the answer they expect is the correct answer. I think that's at least plausible. 3) Only a small section of observed children were interviewed: 28 total, because of a high list of selection criteria, according to the second paper linked in the article. It all feels strangely written.