Forging the digital future | MIT Technology Review

Of course, it worked. He traveled to Cambridge and was drawn to the Artificial Intelligence Laboratory at MIT’s Tech Square, where he worked first on speech recognition and then on to computer vision, then still in its infancy. After receiving his Ph.D., he served simultaneously as a professor of computer science at Cornell University and a researcher at Xerox Palo Alto Research Center, splitting his time between New York and booming Silicon Valley, where he worked on computer vision for the digital transformation of copiers and scanners. Work. “In academia you have more curiosity-driven research projects, whereas in the corporate world you have the opportunity to build things that people actually use,” he said. “My career has been going back and forth between them.”

In the process, Huttenlocher also gained management experience. A long-time board member and eventual chairman of the MacArthur Foundation, he also helped found Cornell Tech, the university’s graduate school for business, law, and technology in New York City, and served as its first dean and associate academic provost long. When Stephen Schwarzman, chief executive of the investment firm Blackstone Group, donated $350 million to MIT in 2018 to establish a computer institute, he was eager to return to lead the school. “MIT has made a bold commitment to be a broad-based leader in the AI-driven era, and the fact that it is reaching all of its schools is exciting,” he said.

The Schwarzman Institute is a working group of more than 100 MIT faculty and staff. By the fall of 2019, a plan had been finalized with Huttenlocher as director and EECS director Ozdaglar named associate dean for academics. “I never believed that everyone wanted to do computer science at MIT,” she said. “Students come here with great enthusiasm, and it is our responsibility to educate these bilingual students so that they are both proficient in their subjects and able to take advantage of these advanced computing frontiers.”

Ozdaglar’s ​​background is in using machine learning to optimize communications, transportation and control systems. Recently, she has become interested in applying machine learning algorithms to social media, studying how the choices people make when sharing content affect the information and misinformation recommended to them. This work builds on her long history of interdisciplinary collaborations in the social sciences, including with her husband, economics professor (and recent Nobel Prize winner) Daron Acemoglu. “I feel strongly that to really solve the important problems in society, these old departmental or disciplinary silos are no longer enough,” she said. “This academy allows me to work more broadly at MIT and share everything I’ve learned.”

Ozdaglar has been a driving force in recruiting faculty to the college, working with 18 departments and bringing in dozens of academics at the forefront of computing. In some ways, integrating new staff into existing disciplines is a challenge, she said. “We have to continue to teach what we’ve been teaching for decades or centuries, so change is hard and slow,” she said. But she also noticed a palpable excitement about the new tool. The school has brought on more than 30 new faculty members in four broad areas: climate and computing; human and natural intelligence; humanities and social sciences; and artificial intelligence for scientific discovery. In each case, they are given an academic home in another department and an appointment within the college, often with laboratory space.

A commitment to interdisciplinary working is integrated into every aspect of the new headquarters. “Most buildings at MIT feel quite homogenous,” Huttenlocher said, leading the way along brightly lit corridors and common spaces with huge glass walls overlooking Vassar Street. “We wanted to make this feel as open and accessible as possible.” While the institute’s high-end computing occurs primarily at a large computing center in Holyoke, about 90 miles west of Massachusetts, the building is comprised of labs and Consisting of communal workspaces, all made of glass and natural blond wood, the space is bright and airy. Along the hall, open doorways offer glimpses of things like giant robots hanging from the ceiling with wires wrapped around them.

Labs and office spaces for faculty research groups working on related topics (possibly from CSAIL and LIDS) are scattered on the same floor to encourage interaction and collaboration. “This is great because it creates connections across labs,” Huttenlocher said. “Even the conference room is not part of the lab or the academy, so people actually have to collaborate to use it.” Another dedicated space can be requested for six months at a time for special collaborative projects. Last spring, the first group to use it focused on applying computing to climate challenges. To ensure that the building is also accessible to undergraduate students, it houses a classroom and a 250-seat lecture hall, where classic Course 6 courses (such as Introduction to Machine Learning) are currently held, as well as new multidisciplinary courses. The soaring central hall is lined with comfortable booths and modular furniture, ready for study sessions.

For some new teachers, working in the academy is a welcome change from previous academic experiences, where they often felt trapped between subjects. “When I started my PhD in 2015, the intersection of climate sustainability and artificial intelligence was just emerging,” said Sherrie Wang, an assistant professor in the Institute of Mechanical Engineering and Data, Systems and Society, where she is also a principal investigator. . When she enters the job market in 2022, it’s still unclear which sector she will work in. practice. “It’s nice to have a group of people with similar philosophical motivations applying these tools to real-world problems,” she said. “At the same time, we are also advancing the development of these tools.”