By The Editorial Team
What does it take to found a successful AI company? Radical Ventures Managing Partner and Co-Founder Jordan Jacobs explored this question with AI roboticist and multi-time founder Pieter Abbeel in the kick-off to this year’s AI Founders Master Class. Over 550 AI researchers from around the globe have registered for the four-part program which was created by Radical Ventures to inspire the next generation of technology-driven company leaders (learn more here).
Pieter is a Professor at UC Berkeley, Director of the Berkeley Robot Learning Lab and previously founded the AI-assisted grading tool company, Gradescope. He is currently the Co-Founder and President of Radical Ventures portfolio company Covariant – a company that has built ‘Covariant Brain’ a universal AI brain platform to make robots smart.
The following excerpt from his conversation with Jordan is edited for length and clarity.
Jordan Jacobs: The central question that we are trying to help people answer for themselves is, “Am I a founder?” And so in both cases (Gradescope and Covariant) you had the sense that there was a real company to be made. In one case from a product and the other from technology. Did you ever sit down and say to yourself, “am I a founder?” What analysis did you undertake to decide that you wanted to be a founder instead of just the professor helping your students take out the technology and commercializing it?
Pieter Abbeel: It comes down to asking yourself, “what am I most uniquely qualified to do?” Because that will give you more satisfaction in life than if you do something that someone else could have done if you hadn’t. In both of these cases, that was the biggest driving factor.
For Gradescope, the unique qualification caught me by surprise. Because it wasn’t a technical or a scientific qualification. It was me, personality-wise, being process-driven; always trying to optimize everything that I’m doing. Always trying to eliminate the inefficiencies. As a professor, I was also a natural user of the product (a grading tool) and I had many friends and colleagues who would also be natural users. It was the drive for optimization and the many connections to the adopters of the product that made me feel pretty uniquely qualified to actually build that out.
For Covariant, it was over two decades of working in the robotics space and writing about robot learning that made me feel 1) very comfortable in the field and confident that I’m really at the frontier and can do something pretty unique; and, 2) extremely familiar with other people who are doing great things in this space, including many of my former students. I knew that we would get along well and so I knew it would be possible to build out a strong team that can drive this forward over time together.
If you are an AI researcher interested in entrepreneurship, the AI Founders Master Class offers participants an opportunity to join conversations with AI pioneers and access to practical seminars and resources designed to support entrepreneurs looking to commercialize their research. AI Founders is hosted on Slack and Zoom with hybrid in-person and video sessions. Next week’s module features computer vision pioneer Fei-Fei Li and explores the imperative of building technologies that solve a customer problem.