University of Toronto associate professor Sanja Fidler is leaving her role as Nvidia’s VP of AI research after eight years at the world’s most valuable company.
**The news: **Fidler announced her departure from Nvidia in a LinkedIn post on Friday.
Fidler, who specializes in computer vision, joined Nvidia as its director of AI at the invitation of CEO Jensen Huang in 2018. She led the company’s AI research lab in Toronto, developing underlying technologies that enable AI systems to perceive, model, and interact with the physical world.
From the source: “The mission of the team was to develop new technologies for simulation to power the creation and modeling of virtual worlds,” Fidler said in her LinkedIn post. “I couldn’t be prouder of what we have achieved.”
Fidler did not disclose what was next for her, but said world models (which are meant to simulate the physical world) are “where the next breakthrough lies, and it is around the corner.”
“I’m super excited for what’s to come,” she added.
The context: Fidler was hired as employee number one at Nvidia’s Toronto AI research lab, now the Spatial Intelligence Lab. She went on to grow the team to just over 80 members, according to its website. Fidler is also a co-founder of Toronto-based AI research organization Vector Institute, which she has said was created to slow Canadian brain drain.
**Final thought: **Fidler’s LinkedIn post noted multiple research milestones Nvidia achieved with world models, including recently demonstrating real-time interactive driving in a fully AI-generated world. She also delivered a keynote presentation on world models at Nvidia’s GTC conference this past March, specifically on their application in autonomous vehicles.
Fidler isn’t the only Canadian AI pioneer interested in world models as the next frontier. The co-founders behind Maluuba, a Canadian deep learning lab acquired by Microsoft in 2017, recently founded Skyfall AI with the intention of using world models to replace nearly all of the operational infrastructure of companies.
Feature image courtesy Toronto Tech Week.