“5 Questions” is a series by the School of Art that asks alumni who are transforming art, culture, and technology about their current work and time at Carnegie Mellon.
Doug Fritz is a creative technologist with a keen interest in systems architecture and a passion for using technology to solve pressing real-world challenges.
After graduating with a BFA in art and BS in computer science, Doug earned his MS from MIT. He had the opportunity to work with Yahoo Research as well as with the team that developed Siri and has worked for Google in many different roles. He now works as Head of Visualization for DeepMind (a subsidiary of Alphabet Inc.), a company focused on pushing the boundaries of artificial intelligence and developing programs that can learn to solve any complex problem without needing to be taught how.
1. How did your art education lead you to, or help prepare you for, your work at DeepMind?
I was always drawn to art as a way of seeing the world. I was lucky to be able to pursue both a BFA in Fine Art and a BS in Computer Science at CMU, and I think it was in their intersection that I found many of the skills and passions that have put me on my current path. Critical theory, interaction, and design all interplay well with building and programming as mediums of creation and understanding.
2. DeepMind is committed to the use of artificial intelligence for positive impact. On the other side of coin, others may see AI as simply a tool to maximize corporate profits or even as a tool for nefarious purposes. What do you see as the ethical responsibilities of researching AI?
Ethical responsibility in this field is incredibly important, and one of the major reasons I chose to work at DeepMind. I am also very proud of Google PAIR – People and AI research effort, and have many colleagues there as well. It’s still very early days for AI research and decisions about how to use it are not for anyone to unilaterally decide. These are decisions for society as a whole, which is why it is important to engage with academics, policy makers, and the wider public. To do so, DeepMind has a technical AI Safety team lead by our co-founder Shane Legg, a dedicated Ethics & Society team working with external experts, and was also a founding member of The Partnership on AI to work with other companies and nonprofits. It is a complex and important task and takes this sort of multifaceted approach and dedication to its prioritization.
As you put it in your question, machine learning techniques are tools. As such, they will likely be used throughout society to find patterns in data and solve problems. Before we can agree on what it means to use these tools ethically, society needs to come together to agree upon a shared set of values that drive ethical responsibility. I would like to see more researchers from fields like the humanities and arts taking computer science classes and vice versa so that they may have a shared vocabulary to enter into the deep technical and societal conversations needed on this topic.
I also challenge more artists to the pursuit of “Critical Optimism” utilizing extrapolative thinking like we see in science fiction novels and artwork as forms that galvanize us to envision a future we can create together. It is easy to be negative, but art has the rare ability to paint a picture of a world that inspires us; and toward which we can strive to build. First, however, we must choose what kind of world we want, which is often the hardest step.
3. Could you speak about one project that you’ve worked on at DeepMind that sticks out for its positive impact or its potential to make a positive impact?
It is hard to pick just one, as I feel like all of the work has potential for positive impact. Personally, I pursue explanations through interactive play. What does an agent do and why. My focus at the moment is on accelerating researchers and helping to understand how agents are solving tasks, which has potential in the longer-term for positive impact. In the short term, I would point to my colleagues in the Applied side of DeepMind, who have made particularly wonderful advances in Health and Energy. Their work is literally helping to save lives and the planet.
4. Could you talk a bit about your time as an undergraduate at CMU’s School of Art? Are there any experiences you had as a student that stand out?
Fundamentals of Computational Visual Form with Golan Levin was a particularly transformative class. Processing, an IDE and programming language created for artists and designers had just started to spread in the community and was first introduced in this class. It opened up an ability to see the intersection of art and computer science. Learning processing acted as a nexus which spread into being able to integrate the teachings of all of the wonderful professors at the school, whom I would like to thank.
5. Do you have any advice to share with our students?
Take advantage of the multidisciplinary nature CMU has to offer, contrary to your beliefs going in, most of the graduates will not be Artists with a capital A as their career, that isn’t a bad thing, art is a critical element to thinking, making, building, and having impact in our society. There are so many opportunities at CMU, it is important to remain open to them, but they aren’t always the ones you expect, which reminds me of the birthday paradox as a lens for viewing opportunity. At the end of the day, it is important to think broadly of how you want to use your precious time at CMU and what you plan to do after you graduate that might not be painting canvases like Rembrandt as I thought it would be when I was a freshman.