32 million battery materials screened in 80 hours ⚡ Microsoft and PNNL fed an AI every known battery

Añadido: 2026-05-16

[00:00:00]32 million battery materials screened in 80 hours.

[00:00:08]Finding a new battery material normally takes decades. Scientists test candidates one at a time. It's slow and expensive. Microsoft and the Pacific Northwest National Laboratory fed an AI model every known battery compound and asked it to find better solid-state electrolytes.

[00:00:26]It screened 32 million candidates in 80 hours. That would have taken traditional research over a century.

[00:00:33]The AI found a material that uses 70% less lithium. Lithium is the bottleneck for every electric car and phone battery on Earth. This single discovery could change the economics of the entire energy transition.

[00:00:47]Follow for more AI facts about the future being built right now.