CoreWeave: Physical AI Will Help Us Travel Safely, Not Kill Us All

There has been a lot of concern recently about the effects AI could have on humanity. One 2017 Nobel peace prize winner even suggested AI might kill us all. It’s not just fearmongering, either. On September 16, Reuters reported that OpenAI would regularly disclose unexpected or unauthorized model behavior, amid renewed concerns about increasingly autonomous systems. But the debate over whether AI could threaten our continued existence can obscure work already helping people understand and improve the physical world. One company spearheading this area is the section of CoreWeave that was formed out of the latter’s acquisition of Monolith AI. I talked to the company about how its physical AI is helping not just mobility, but many other areas besides.

Monolith AI built its business by helping engineers learn from test data. Automotive was a central focus, particularly enhancing the process for assessing new battery chemistries. CoreWeave acquired the company in October 2025. Its team and methods now underpin CoreWeave’s Physical AI Field Engineering service, launched September 10. “Monolith was strong in automotive and at the intersection with hardware,” says Richard Ahlfeld, now CoreWeave’s senior vice president of physical AI. “We weren’t a supplier for autonomous vehicles, but we were in that space.” The service embeds specialists with customers to build and validate models using their own engineering data. CoreWeave says its approach has been applied across more than 100 projects in automotive, aerospace and robotics. The ambition extends well beyond faster cars.

CoreWeave Physical AI Moves Beyond The Buzzword

How CoreWeave Helps The Humans On The Pit Wall

MORE FOR YOU