ML screens 50M compounds, all 7 bee repellents work in field tests without harming bees

VraserX · x · 2026-09-28

UC Riverside researchers used machine learning to screen over 50 million compounds for bee repellents; all seven candidates tested in the field repelled bees from honeycomb without harming them.

The goal is keeping bees away from pesticide-treated crops. The author argues blanket AI opposition on environmental grounds is lazy — this is still research, but a concrete example of AI delivering environmental benefit beyond its electricity footprint.

Original post →

More from Research

Research channel →