Billionaire Stephen Schwarzman’s mega gift aims to develop ethically grounded AI leaders and accelerate America’s competitiveness.
In 2018, ChatGPT was four years away and artificial intelligence was a wonky concern for tech and policy thinkers. But thanks to what billionaire financier Stephen Schwarzman saw on business trips to China, he was already worried about Chinese dominance of the technology and wanted to increase American competitiveness. He was also concerned about the potentially disastrous impact on the U.S. work force, economy, and national security. That year he gave the Massachusetts Institute of Technology $350 million to develop campuswide interdisciplinary programs to help its students better understand AI’s impact on people’s lives.
The goal: to responsibly harness the power of AI by shaping a more ethical and thoughtful generation of AI leaders who will help U.S. companies dominate the field.
“The wonderful parts of the technology will probably turn out to be more wonderful than people imagined,” Schwarzman told the Chronicle. “But on the negative side, we can’t just let things happen because the technology would be too powerful to suffer the consequences of that.”
In the world of AI, that gift was a lifetime ago. While MIT has been getting its program underway, AI products have been rolled out to consumers, and nearly 60 percent of Americans believe AI poses a high risk to society, according to a Pew Research Center study. Fears about job losses are growing.
Schwarzman’s donation launched MIT’s Schwarzman College of Computing and was the lead gift toward the college’s goal, now completed, to raise $1 billion to establish the college and to integrate programs in computer science, artificial intelligence, and related fields across MIT’s academic departments. New interdisciplinary programs have been created to prepare students for an AI-saturated world and help them understand the social and ethical implications of digital technologies. The university also hopes to encourage collaboration between government and industry to ensure new technologies are deployed in more socially responsible ways.
Schwarzman is not an MIT alum. He previously gave smaller AI-related gifts to Harvard, where he earned his MBA, and Oxford but says MIT was the natural choice for his largest AI gift.
“From my perspective, they have the best-known brand name in the world for technology,” he says. “Almost everyone I would meet [while] learning about AI either went to MIT or were taught by somebody who went to MIT.”
Tech and philanthropy experts agree that a counterbalance to the AI industry is necessary. Philanthropy directed to academic programs and research institutions can support long-term independent work on AI safety, ethics, and social responsibility. While academia does not move at the speed of technology, it can have deep, long-term impact through research and influence industry leaders for decades to come.
“The number one thing academia does is it establishes independent research credibility outside of what the companies are telling you,” says Devin Kim, a former employee of Elon Musk’s xAI, who is now president of the nonprofit Center for AI Safety. “Universities are able to interpret the research and speak freely about what the implications are without being bound by short-term commercial interests.”
Reconceiving Education in the Age of AI
MIT realized that effective education in the age of AI has to look different than it has in the past. Traditional siloing of expertise won’t work when AI is expected to touch nearly every part of people’s lives and is changing the way people in disciplines outside of computing are advancing their work, says Daniel Huttenlocher, dean of the Schwarzman College.
Dave Burk
MIT’s Schwarzman College of Computing has new interdisciplinary programs to prepare students for an AI-saturated world and help them understand the social and ethical implications of digital technologies.
The university has used the gift to expand the Schwarzman College’s computing faculty to better connect computing with other disciplines, says Huttenlocher. The gift has also helped MIT make ethics in computing an academic priority.
MIT has since hired nearly 50 new Schwarzman College faculty, including 29 who specialize in computer science and artificial intelligence and 20 new “shared” faculty whose expertise lies in both computing and other disciplines across the humanities and science. They span 19 departments across the university and joint posts in the Schwarzman College, creating a level of cross-pollination that didn’t previously exist.
The gift has funded a new undergraduate major in AI and decision making that knits together multiple disciplines, and the money has helped pay for the college’s new building, where computing, AI, the new shared faculty, and graduate students all have space, making it easier for them to collaborate.
Such cross-pollination is expanding the use of AI across all disciplines, and Huttenlocher says it will improve how technologists build, fine tune, and iterate AI tools going forward.
With this new approach, he says, experts in many fields can gain a deeper understanding of AI so they can better harness it. Meanwhile, computing experts are gaining greater exposure to the work of their counterparts in other fields. That exposure is giving the technologists a better understanding of how to create and train AI tools to better serve others.
Schwarzman’s donation also launched Social and Ethical Responsibilities of Computing, an institute, known as SERC, that seeks to develop what the university calls “responsible habits of mind and action” and foster a culture that prioritizes creating technologies that are in the public’s best interest.
“It’s very difficult to teach morals. Just teaching an ethics class on its own is probably not going to work,” says Agustin Rayo, dean of MIT’s School of Humanities, Arts, and Social Sciences. “What we need to do is make sure that when students think about these issues that it’s clear that this is something not just the ethicists care about but everyone cares about, including the computer scientists they idolize, because otherwise, it’s not something they’re going to take seriously.”
Philanthropy: One Piece of the Puzzle
To date, more than 300 students across 25 departments and labs have participated in programs offered by SERC. More than two dozen of the classes taught by the “shared faculty” are now offered, and at least 1,000 MIT students are taking the courses annually.
Within the institute, academics from across computing, data sciences, humanities, arts, and social sciences are developing new teaching materials to incorporate into existing courses, and they have created classes taught jointly between computing and humanities or social science faculty.
Schwarzman is pleased that so many students are now participating in these programs. He says that was part of his “dream” when he decided to give MIT the money.
One of the new courses examines the ethics of computing and is taught by two top academics — the philosopher Bradford Skow and the computer scientist Armando Solar-Lezama. It is focused on providing students with a thorough understanding of ethical issues like privacy, algorithmic bias, disinformation, and other factors that arise with powerful new technologies. Because students are learning about these issues from respected experts, they are more likely to engage with the material, says Rayo.
“We wanted the quality of the instructors to convey how important these issues are,” he says. “This class is focused on generating deep understanding of the issues so we don’t end up with students who are ideological, so we instead end up with students who have the tools to understand the complex issues they’re going to face and then use their own moral compass to decide how to do it.”
With such enormous sums of money being spent to make AI systems more capable, Kim, with the Center for AI Safety, says efforts like MIT’s are necessary even if quantifiable changes can be slow to materialize. Little money is being spent to understand the reliability and societal impact of AI. For-profit companies are not going to play that role — philanthropy must step in, and universities such as MIT can provide an independent and trusted perspective, he says.
Efforts like MIT’s have the potential to be very effective, says Rob Reich, a Stanford University professor who specializes in science and AI ethics and who advised the U.S. Artificial Intelligence Safety Institute, a government entity that has since been renamed and today works with industry to facilitate research on the potential of AI systems. Many prior efforts to get academia to work on questions surrounding AI have been stymied, he says, by tech companies that can afford to lure university researchers away with high salaries and access to better computing infrastructure and top-quality labs.
MIT’s interdisciplinary structure provides an advantage that big tech companies do not have: a stable of professors with diverse expertise working together across departments to improve the thinking and motivations of fledgling and future technologists.
Philanthropy isn’t the only solution to the problems surrounding AI, but it’s one piece of the puzzle, says Reich, who was co-director of Stanford’s Center on Philanthropy and Civil Society. While he would like to see more federal government funding going to such efforts, he doesn’t believe that will happen in the near term given the harsh partisan divides on Capitol Hill.
“My defense of big philanthropy is when philanthropy’s working well, it’s one way to have a long-term effect on something like this,” says Reich. “Giving to universities for these kinds of programs is one potential good use of philanthropy.”
Well-Rounded Tech Leaders
Universities can’t do everything on their own, says Daniel Barcay, a former software engineer at Google who is executive director of the Center for Humane Technology, a nonprofit watchdog organization. The market incentives that drive the creation and design of new technologies must shift.
Most technologists want to build tools and systems that benefit society, but they end up trapped in a market structure driven by a winner-take-all mentality, Barcay says. Facing intense competitive pressure, AI leaders rush to create the next iteration of their technology before their competitor beats them to market, regardless of its impact on society.
Real change will come when the incentives change. “We can do better,” he says. “But it requires not just us thinking about ethics more deeply, although we need to do that, it requires us having the right laws that stop the races to the bottom.”
MIT is not blind to the need for more than just ethical leadership. Asu Ozdaglar, head of MIT’s department of electrical engineering and computer science, says the technology industry, academe, and state and federal governments must work together on developing the future of AI and other new technologies so they will benefit society. The public and industry-facing programs MIT is developing will help make such collaboration more common, she says. To that end, MIT leaders are working to bring together lawmakers, industry leaders, and academic researchers to develop approaches for more socially responsible technology regulations.
“It’s important to educate scholars to incorporate [ethics and social responsibility] into our research and to work in partnerships with industry and government to really understand how we would like to develop this technology so that it’s beneficial for our society,” Ozdaglar says.
It is too soon to tell whether these efforts at MIT will eliminate the “move fast and break things” ethos that led to some of the more disastrous outcomes of today’s social media landscape. Kim says that while some of those who are building AI systems are thoughtful about the societal implications of their work, the “vast majority” are not. He thinks university efforts like MIT’s might be one way to change that.
MIT’s programs can contribute to producing more well-rounded tech leaders who have an interdisciplinary mindset, says Kim. “At the end of the day, if the researchers and employees of these labs start to feel their work is doing a lot of harm and they’re having more reservations about the work they’re doing, that can impact the decision-making of their companies and leadership,” he says. “I do believe these educational programs help move the conversation along in these critical spaces.”
Reporting for this article was underwritten by a Lilly Endowment grant to enhance public understanding of philanthropy. The Chronicle is solely responsible for the content. See more about the Chronicle, the grant, how our foundation-supported journalism works, and our gift-acceptance policy.
