Carlos Guestrin, the Fortinet Founders Professor of Computer Science at Stanford University, will lead the Stanford AI Lab.
As artificial intelligence rapidly transforms our world, its early breakthroughs were pioneered at Stanford University. In 1963, the esteemed Professor John McCarthy founded the Stanford Artificial Intelligence Lab (SAIL), which over the years became a hub for groundbreaking innovations in fields like computer vision, bioinformatics, game theory, information retrieval, robotics, natural language processing, neural networks, and more—many of which serve as the foundation of AI today.
Now, SAIL is teaming up with the Stanford Institute for Human-Centered AI (HAI) to continue to build a powerful AI research hub on campus. This collaboration will drive the development of cutting-edge AI technologies and their real-world applications, ensuring ethical guidelines are in place to ensure the benefits of AI are accessible to everyone.
In this new phase, Carlos Guestrin, Fortinet Founders Professor of Computer Science at Stanford University, will take on the role of director of SAIL. A leader in machine learning and its applications, Guestrin has held leadership roles at multiple AI startups, led Apple’s machine learning team, and founded Turi, a platform for building intelligent applications.
Guestrin succeeds Christopher Manning, the Thomas M. Siebel Professor in Machine Learning in the Departments of Linguistics and Computer Science at Stanford University and an expert in natural language processing, who led SAIL through a period of significant growth since 2018.
In the following discussion, Guestrin shares his vision for SAIL, reflects on the complementary collaboration with HAI, and speaks to the future of AI scholarship at Stanford.
What excites you about leading SAIL?
SAIL is a historical organization that has had tremendous impact in thought leadership and innovation over the last six decades. Through every wave of AI that we've had, the different techniques and applications and the recent tremendous growth of AI, SAIL has been at the forefront of this innovation. Just look at the transformational impact of AI that has happened over the last five years — many of the techniques out of this group have been core to those developments.
What is your vision for your tenure as director?
SAIL is the Stanford center of excellence for AI technology, and I think that that's what we need to continue to invest in. My goal is to support the community, the students, faculty, and staff. They're enabling this next generation of AI and AI leaders into the world. We want to make sure that we have a welcoming and collaborative environment that fosters this innovative spirit. We want to support junior faculty in their growth as they bring in new ideas and energy into the group. And we want to continue to get the ideas from this group out into the world in the most impactful way. So really, my goal is to shepherd the group to continue its mission.
How do you anticipate SAIL and HAI’s integration changing the work these organizations do?
It's an exciting direction. I love HAI’s mission and ongoing impact, its broad vision, work in policy and ethical considerations, and I think HAI’s work is complementary to what we've done in SAIL. In fact, SAIL has been around much longer than HAI, and many of the core people who started HAI came from SAIL, like HAI Associate Director Chris Manning and HAI Co-Director Fei-Fei Li. A large fraction of the core SAIL faculty, some 40-plus members, are involved in HAI in some way, either as an active member or as somebody whose research has been funded by HAI.
In a sense, it's not a big change, but a recognition of the ongoing collaboration from a practical perspective. Together, we can help double down and accelerate each other's efforts.
SAIL will become a center within HAI like the Center for Research on Foundation Models or the Stanford Digital Economy Lab, but we will continue to independently drive our research and education goals and continue to be Stanford’s center for excellence for AI technology. And in that sense, my view is that SAIL is HAI’s arm or partner in technical innovation in AI. We're also looking forward to leveraging resources that HAI has, including getting the word out about our work through communications, events, development, and more. The collaboration will support and enhance the impact of SAIL’s faculty and students.
What are you most excited about right now in your own work in technical AI?
I'm really interested in two sides of how AI is impacting the world right now. So one side is how it's being created, and the other side is how it's being used. On the creation side, we want to empower developers to create trustworthy and scalable AI applications. In a way, this latest technology has enabled more and more people to build AI, but we also haven't had the foundations to do that in a scalable, trustworthy way. But on the other end, I'm also interested in how we as humans — how you and I — consume the answers from an AI system. For example, if I ask a question about AI to ChatGPT, I have a pretty good sense of whether it's correct or not. But if I ask a question about astrophysics, when the system responds to me with complete confidence, I have no idea what I'm consuming. So what does it mean for us to trust, verify, and depend on the reliability of this AI system? So, on both sides, our goal is to enable AI applications to be trustworthy and scalable.
What does the future of AI research and education look like at Stanford?
The future is bright. We have colleagues who are developing the core AI technologies, things like new methods and algorithms that will lead to large-scale innovations in AI. We have scholars working on new frontiers of AI, areas like robotics and computer vision, where there is still a lot more that we can do. And we have people working on applications, be it in drug discovery, biology, social sciences, and education. Any corner of the university where AI is making an impact or there's a seed of transformation, we’d like SAIL to be a partner for developing the core technical innovations to support that.