Facebook partners with Carnegie Mellon University to use AI in a bid to find new “electrocatalysts”, which help store electricity from renewable energy sources
Context & Ripple Effects
This partnership extends a pattern rather than starting one: Facebook has been building academic AI ties since its partnership with 17 universities including MIT, Stanford, and Harvard in 2016, and opened a dedicated AI lab in Pittsburgh in 2018 — Carnegie Mellon's home city, making this a natural deepening of an existing footprint.
It also slots into Facebook's energy agenda: Yann LeCun's team was already designing its own energy-efficient chips for video processing, and the company later declared it had reached its goal of powering global operations entirely on renewables after $8B in green investments. Storage is the missing piece of that story, which is what electrocatalyst discovery targets.
First-order effects
- Carnegie Mellon researchers gain access to Facebook's AI tooling and compute to accelerate electrocatalyst screening, while Facebook gets a research pipeline aimed at storing the renewable electricity its operations run on.
Second-order effects
- Rival cloud and platform companies face pressure to show comparable materials-science and storage R&D, since Facebook can now pair its renewable-energy claims with a proprietary path to solving intermittency.
Third-order effects
- If the pattern holds, large tech firms shift from being energy consumers with procurement pledges to being energy R&D actors — using AI-driven materials discovery to shape the storage market their datacenters depend on.
The trend: Big tech is pointing its AI capability inward at infrastructure problems, moving from buying renewable power to researching how to store it.