OSTP Director Arati Prabhakar on Harnessing AI to Achieve America’s Great Ambitions
As prepared for delivery at AI Aspirations: R&D for Public Missions, a White House conference.
Welcome, everyone!
I am delighted to welcome you here today for AI Aspirations.
We have a wonderful group in the Bloomberg Center today! Raise your hand if you’re here from a university or nonprofit. From a company. From government. Thank you all for joining us.
An exceptional team from across government has put this conference together. We’ll be welcoming some terrific people to the stage throughout the day.
So, let’s start with what this conference is all about.
President Biden likes to say, “America is the only nation in the world that can be defined by one word: possibilities.”
And that’s exactly why we are here today: to explore the possibilities ahead if we put artificial intelligence to work in powerful and responsible ways to achieve big things for our country.
We’re here today to look into the future with a purpose; to imagine the future so we can build it.
When the President and Vice President set out last year to address what they call “the most consequential technology of our time,” they were clear that we have to put AI on the right track for the American people. And that means managing AI’s risks so that we can seize its benefits.
It’s a priority for this Administration. That’s why we have delivered: The first-ever voluntary commitments from AI companies. A landmark executive order on AI, pulling every lever under existing U.S. law.
It’s why we’re working with allies and partners around the world and we continue to work with Congress on a bipartisan basis for legislation.
That’s an important start on the journey to get AI right.
And the reason to do all the work of managing AI’s risks is so we can use it to achieve our great ambitions. The reason to build a stable platform is to stand on it to reach for the stars.
And that’s what today is all about.
We know that companies and investors are moving out to use AI to serve other companies and consumers. But AI can help us achieve our country’s great ambitions too.
Because here we are, a quarter of the way into the 21st century, and America’s ambitions are as great as they have ever been: health and opportunity for every person; meeting the climate crisis and rebuilding our infrastructure; prevailing in a global competition that is growing fiercer every day, so our people can thrive in peace.
President Biden has taken historic action on each of these immense challenges. He is investing in America with landmark legislation that’s rebuilding roads and bridges, deploying clean energy at a massive scale, bringing leading-edge semiconductor manufacturing back in the United States, and preventing disease—and making health care easier to access and easier to afford. And with the Cancer Moonshot—doing the work to end cancer as we know it.
This is extraordinary progress that pushes the fruits of earlier R&D out into the world, deploys them, scales them, and makes sure they reach all Americans. That’s how we win today.
And when we look to tomorrow, to win the future we will need many more new advances from R&D.
And so our focus here today is how R&D—supercharged with today’s AI—can transform public missions for the future. And what it will take to make that happen.
That’s why we’re here. Let me tell you what we’ll be doing throughout the day. Today’s conference has several elements.
One is a collection of visions of what can be possible if we go big with AI. We’ll hear from seven amazing people from very different parts of government.
Each one will share an AI Aspiration, an example of the transformative national capabilities that could be realized with AI. First in a short talk here on the main stage, and then in a breakout session for a deeper dive.
Across these seven Aspirations, we’ll see a wide range of different AI technologies:
AI models trained on different data types. Some will build on language models and other commercial models, but almost all of these Aspirations require their own AI models. For example, models trained on molecular data and medical data for new medicines. Trained on thermal and electronic properties for sustainable materials. Trained on barometric readings and radar patterns for weather. Trained on voltages and currents for managing an electrical grid.
We’ll also see what it takes to make these big aspirations happen. As anyone who works with AI knows, it’s much more than just showing a bunch of data to a model, and sitting back to watch the magic happen. You have to contend with all kinds of hurdles.
Often the data you need is locked up because it’s proprietary or sensitive. Sometimes it doesn’t even exist yet. Once you train or fine tune a model and build the system around it, you have to validate it in the real world. And once you’ve proved your new approach works, you still have to change minds to start implementation, and to fully scale. All of that is what it takes to change people’s lives, which is the point of the exercise.
And one more thing we’ll see in these seven AI Aspirations: None of them can be accomplished by industry alone. Many require deep research that’s beyond the reach of corporate product development. Some require training AI models on data from more than one company. All require the deliberate focus to deliver on a public mission, which is different than a market opportunity. Achieving these aspirations will take government working with universities and labs and with companies, and workers, and communities.
To build the future, we will need to do together what we can’t do separately.
Those are some of the common themes you’ll see across the seven Aspirations.
That’s one element of today’s meeting.
The next element is a collection of demos, 11 demos that are just a small sampling of work that is already underway today, from a host of different federal R&D organizations.
Here again, you’ll see the variety of AI technologies for these very different applications. And you’ll see some of the specific advances that are already coming into our lives today and are paving the way for even bigger things ahead.
The next element of the day is about risks. This is about grappling with a question that is integral to the advancement of any powerful technology. That question is: “What could go wrong?” It’s a question we ask because it points the way to building in protections and mitigations before a new technology is deployed.
Each of the seven champions for Aspirations will address the specific societal risks for their area and we will also have a terrific panel discussion with experts who will share their perspectives on risks, as well as practical ideas for getting it right from the start.
So, that’s what you’ll hear from us today: visions in seven different domains, demos of real progress from research today, what could go wrong, and how to deal with it from the start.
The final element of the day is from you. All through the day, over lunch and at breaks, you’ll have time to talk with each other and brew up some big ideas of your own. So you’ll find Post-its on whiteboards to invite you to share your AI Aspirations for America.
What is the transformative national capability that you can see ahead with AI? We can’t wait to hear from you—because there is so much more to do.
For this conference, we chose seven civilian and domestic public missions.
Of course, AI is also a gamechanger for our security. A decade ago I was the director at DARPA focused on national security, and even then we were already using AI for everything from self-driving ships, to getting more bandwidth from a fixed slice of electromagnetic spectrum. For security here at home, AI will be essential to boost cybersecurity and to protect our critical infrastructure.
Even in the seven areas we selected—there are abundant other opportunities for AI to make a big contribution beyond the specific applications that you’ll hear about from us.
In energy, we will discuss what’s possible for our electric grid and we can also use AI to manage the energy use in a building, to design the advanced materials for better batteries, and to design whole new energy sources.
For dealing with climate change, we will discuss AI for weather forecasting, and we can also use AI for the tough challenge of reducing agricultural emissions. We can use AI to support the biodiversity of our ocean that sustains all life on Earth, and to explore the courses of action still ahead to fully meet the climate crisis.
For better health outcomes, we will discuss AI for creating new medicines faster and we can also use AI for better diagnostics and mental health & clinical support. We can use it to boost food security and nutrition, to improve public health & prevention—to avoid disease in the first place
We will discuss AI for safer transportation infrastructure. And we can also use AI for increasingly autonomous vehicles, and better traffic flows and transit that works for more and more people.
And similarly, there are many more possibilities ahead for materials, education, training, and government services than the Aspirations we will present today.
And underpinning all of these applications, we can use AI to change research itself. Actually changing how we expand human knowledge.
So today is a call to action. A call to universities and companies; labor and civil society and philanthropists; to federal agencies and Tribal, state, and local governments; and to Congress.
The prospects ahead of us are exhilarating and the work ahead of us is daunting.
We must do this work—and we can do this work.
Every country is racing to use AI to build a future that embodies its own values. And whatever else we may disagree about, none of us wants to live in a world driven by technology that is shaped by authoritarian regimes.
So, President Biden and Vice President Harris have been clear: American leadership in the world today requires American leadership in AI.
That is why we must to this work. And here’s why I know we can do this work.
America is built on the idea of striving for a better future. This is a country where we come together to do excruciatingly hard things. And when we stumble, we get up and try again and, ultimately, we accomplish seemingly impossible goals.
That’s how America does big things.
President Biden closed his State of the Union address by saying, “Let’s build the future together.”
It’s hard to imagine a more exciting invitation. So let’s get to it. The future is waiting.
To learn more, visit AI.gov/aspirations
###
