Generative AI: Shaping the Future

Good morning! I’m delighted to see such a packed house – and hello to all the alumni on the livestream!

I want to thank Daniela Rus, Cynthia Breazeal and Sertac Karaman for making this amazing symposium happen. And thank you all for being here to help kick off Day One of MIT’s Generative AI Week.

In this gathering of generative AI experts, I want to be very clear that I am not one! But one of my favorite things about the MIT community is the incredible spirit of openness and intellectual generosity.

I happen to be a cell biologist. But if I’m curious about how gen AI might affect my discipline – or any other discipline! – I just start asking questions, and very quickly I’ll find myself in conversation with some of the most brilliant, innovative people in the field.

Today’s event, and the symposiums to follow, are designed to showcase that brilliance and innovation. We have a global audience today – and, as always, we’re delighted to share our work with the world.

But I also see this event as a crucial opportunity for crosspollination, within the Institute. The idea is for everyone here to get a sense of the full spectrum of generative AI topics and questions other people are pursuing on campus.

I know we can count on fascinating panels and speakers today. But to me, the most important measure of success will be the introductions, the conversations and the collaborations that blossom afterward.

So, whatever your particular interest, if you hear something you want to understand more deeply…if you sense some cool new intersection with your own ideas that you haven’t encountered before…or if you feel inspired to invite someone to collaborate with you…please don’t be shy! Ask questions! Introduce yourself! The true purpose of this gathering is to generate creative collisions between people and ideas – in the very best spirit of MIT.

The structure of today’s discussions reflects three crucial realms in which the people of MIT are and have been engaged in this burgeoning field: They’ve helped build gen AI’s intellectual foundations, they’re advancing its applications, and they’re exploring its ethical dimensions and its societal implications and impacts.

As many of you know firsthand, the campus has been buzzing with activity around generative AI for months. For instance,

• This fall, we funded seed grants for 27 faculty projects to accelerate research on how AI will transform people’s lives and work. Because the enthusiasm was so broad and overwhelming, we’ve already called for a second round of proposals.
• The MIT Work of the Future initiative is using real-world data to think about how AI might create better jobs and allow workers to help drive innovation instead of being driven out by it.
• A few weeks ago, at an event called MIT Ignite, students and postdocs offered compelling ideas for AI entrepreneurship – from mental health to drug discovery to immigration and education.
• And through a partnership with Harvard called the Axim Collaborative, MIT is exploring educational aspects of gen AI that could help underserved students reach their potential.

The range and depth of all this new work is fantastic. MIT is building a whole new reputation for itself as the Massachusetts Institute of “the Impact of” Technology!

In that spirit, I want to highlight another area where MIT is working to make a crucial difference: In helping to inform and shape policy regulations around gen AI – regulations that will shape the future for us all, in the US and around the globe.

Working closely with MIT’s Washington office, faculty experts developed a policy paper specifically geared to help leaders in Congress and government agencies understand the norms, regulations and institutions it will take to ensure a future in which we contain the societal risks of generative AI – and make sure that its advances are broadly beneficial.

In this case, the stakes are too high to just “try stuff” and observe the outcomes. We need to make sure that sensible regulation is built in right from the start.

With any new technology, one mark of a successful regulatory regime is that it inspires a productive industrial research agenda. Just as calls to make cars safer inspired automakers to develop seat belts, air bags, lane sensors and anti-lock brakes, we need the commercial players developing generative AI to join us in the project of making sure that it will lead society to exciting new destinations – without driving over a cliff.

In this work, as with ChatGPT itself, the quality of the results we achieve will depend on the quality of the questions we ask.

So, let’s “act as if” we’re a community of intensely creative people with unusual technical insight…charged with helping a late-stage capitalist society plan to safely integrate a rapidly advancing new technology…in ways that are humane and broadly beneficial…that help solve urgent problems…and that usher in an era of shared prosperity. I honestly can’t think of a challenge more closely aligned to MIT’s mission.

It’s a profound responsibility. But I have every confidence that we can face it – if we face it head-on and together.

Thank you.