The title is from the MIT Technology Review. A direct quote from the article: "In fact, a growing body of educators are already arguing that the world's worst-off children are better off being educated by machines."
And the author offers that, "Given the level of math phobia present in American schoolchildren and the sorry state of financing for education, it's worth asking whether or not the trend lines of declining quality in education and increasing quality of educational games have already crossed for a significant portion of American students." And then admits he "made up" the graph to accompany the article. But wait, it gets better. Someone read his post and sent an "improved made up" graph (more wiggles means more realistic?). Am I expecting too much from MIT?