New weekly newsletter filters AI-for-Science papers down to 3-5 transferable ideas

bravo_abad · x · 2026-09-18

Physicist Jorge Bravo-Abad runs Discovery at Scale, a free weekly Substack that filters hundreds of weekly AI-for-Science papers down to 3-5 ideas researchers can actually carry into their own work. Each issue groups findings by what they teach rather than paper by paper, and includes: the concrete ML mechanism behind each result, one number worth remembering, a clear split between what is demonstrated versus speculative, and the lesson that transfers beyond the original field. Aimed at researchers in physics, chemistry, materials science, biology, math and ML.

Original post →

More from Research

Research channel →