Latest meta-analysis on using ai for supplementing instruction

Found a recently published meta-analysis in Nature aggregating effects of using ai to supplement instruction across 51 quasi-experimental studies.

The results look promising, but my biggest takeaway is that all the studies relied solely on existing ai. I think that’s important; we don’t need more when there’s already so much that's already available.

Submitted May 12, 2025 at 04:36AM by variancekills https://ift.tt/lJU1jtu