The research is interesting, the article is... drivel and just pasting internet memes on top of research and cherry picking one mediocre result from some really interesting ones.
Google found nothing of the like. Google trained a neural net to recognize things with no prior knowledge. "Amazingly", among the things it learned to recognize was cats, but not even as well as other things like people. So this whole article just missed the point and tries instead to cash in on an internet meme instead of reporting.
