Skip to main content

Engaging the Public Helps Brian Scassellati Build Better Robots

brown-haired man with glasses and blue shirt, smiling
Brian Scassellati was a 2020-21 AAAS Leshner Fellow working on artificial intelligence.
Photo credit: Courtesy of Brian Scassellati

As a professor of computer science, cognitive science, and mechanical engineering at Yale University, Brian Scassellati gets a lot of calls to lead robot demos for kids. The robots he develops provide support for education and therapy and are often specifically focused on diagnosing or treating autism spectrum disorder, so he also gets a fair amount of media interest in his work. However, during his year as a 2020-21 AAAS Leshner Public Engagement Fellow, Scassellati hoped to devote more time to writing a book geared toward the public. Because of the COVID-19 pandemic, he had to put most of this on hold to manage his lab and students during the crisis. Yet other opportunities arose to engage with public audiences over the past year.

Because so many meetings were held online, Scassellati was able to increase the number of talks he gives to public audiences to about 1-2 per month compared to about that many per year before the pandemic. For example, the Autism Science Foundation’s annual meeting includes a “Day of Learning” for the public, which he spoke at this year, and afterwards received quite a few email follow-ups. Often people will write to ask whether he can research specific questions, and sometimes they will mention something Scassellati and his team hadn’t thought of. For example, they will soon be expanding their work to do more for young adults with autism because they have received so many requests for this.

Scassellati also noted that public engagement is built into his work through inclusive design: their products are tested with end users throughout the development process. And even earlier in the process, they are always careful to be sure they are trying to solve a real problem -- information on this often comes not just from end users, but from other experts who work with them, such as clinical staff. He and his team thus rely on partnerships with institutions and organizations that provide direct services to people with autism or with other groups they are designing tools for.

Opportunities to engage on AI-related policy also emerged during Scassellati’s fellowship. In a typical year, the fellows participate in visits with policymakers on Capitol Hill during their initial orientation, but this year these visits were held virtually at the end of the program. Scassellati led one small group of fellows in preparing a short policy brief focused on the need for better incentives for AI researchers to remain in academia versus going into industry (he notes he is one of only two faculty at Yale qualified to teach their main AI course). Following the briefings, he took the lead in organizing several meetings for the fellows with the National Artificial Intelligence Research Resource task force, a group launched just the month before by the White House Office of Science and Technology Policy (OSTP) and the National Science Foundation.

In part because of the Leshner Leadership Institute’s encouragement to focus on institutional changes that can support public engagement, Scassellati has also been making use of his leadership role with the Association for the Advancement of Artificial Intelligence (AAAI) to reframe how they approach media outreach. They have typically considered themselves a resource to the media, but now he wants to help AAAI do more proactive outreach, and strategically share their own messages about AI.

Scassellati also has become more convinced over the past year that he does indeed want to write a book. Much as he was inspired by the works of neurologist and author Oliver Sacks, he wants to inspire others to go into the field of artificial intelligence. Also, because Scassellati has a background in psychology as well as computer science, he realized he brings a lens to his work that many other researchers in the field do not: he sees AI as a tool for better understanding human behavior (as demonstrated by experiments in which robots are easily able to manipulate people to lie to one another). This perspective on AI research raises important questions about what could, or should, be done with the knowledge that machines can be programmed to change our behavior. Scassellati has written the first chapter of his book and plans to continue making progress over the next year.

The AAAS Leshner Leadership Institute was founded in 2015 and operates through philanthropic gifts in honor of CEO Emeritus Alan I. Leshner. Each year the Institute provides public engagement training and support to 10-15 mid-career scientists from an area of research at the nexus of science and society