El Salvador just ran a pilot in 171 public schools where 1,198 volunteer students scored similarly to the national average results of students in Germany and Sweden in the 2022 PISA test. This is roughly 100 points above El Salvador's national mean (but from a voluntary, non-representative sample). The program is now rolling out nationally, but scaling seems to be the hardest problem in education.

In general, I think we should be way more ambitious than just aiming to improve PISA scores, but perhaps for El Salvador this instrument is a valuable way to measure the progress of its public system.

The pilot appears to be a broad modernization package that includes structured pedagogy, reading/math remediation, principal support, connectivity, devices, and some "curriculum-aligned adaptive AI tutoring" by Grok/xAI.

I want this to succeed, and I want to know how we'd know. If you've worked on scaling school interventions, what breaks first? And what evidence would actually convince you the effect is real? Open discussion welcome.