X @nayibbukele
While we can’t establish direct causality (given all the other improvements), the program’s results are “mind-blowing,” in the OECD’s own words.
And it hardly seems plausible that students would have performed even better than the averages of Sweden and Germany if we had taken AI out of the equation.
So, while we can’t prove causation, we can reasonably infer that the AI tutors had something, or perhaps a lot, to do with these results.
[quoting @1917AndAllThat: "In the interests of fair play, El Salvador schools using AI and personalised learning seem to have done better than comparators. But significantly confounded by major infrastructure upgrades to the participating schools, including running water. https://t.co/rwZuza93JK"]
Notes
While we can’t establish direct causality (given all the other improvements), the program’s results are “mind-blowing,” in the OECD’s own words. And it hardly seems plausible that students would have performed even better than the averages of Sweden and Germany if we had taken AI out of the equation. So, while we can’t prove causation, we can reasonably infer that the AI tutors had something, or perhaps a lot, to do with these results. [quoting @1917AndAllThat: "In the interests of fair play, El Salvador schools using AI and personalised learning seem to have done better than comparators. But significantly confounded by major infrastructure upgrades to the participating schools, including running water. https://t.co/rwZuza93JK"]