National Security Commission on Artificial Intelligence Conference – Senator Chuck Schumer Remarks
Transcript
Thank you very much, Eric, and thank all of you. Those in government and out, who are working on these issues which are so important to both the national security of our country and to the future economy of our country, as well, to sort of combine. And so, I, I thank you for the commission for the good work it has done. I want to thank Eric and Bob Schmidt for getting me to speak here. When Congress set up the commission, I couldn’t think of two better folks to lead it. And so, I very much appreciate that because this is, it is one of the issues that is probably the most important facing us. If you looked at the ratio of how much attention this issue gets, in terms of what government should be doing, as opposed to what’s important. It’s one of the lowest ratios that there is. And, um, that is not good for the future of the country. We all know that artificial intelligence is changing everything. How we communicate. How we conduct commerce. How we protect and store and analyze data. How we protect ourselves. It’s going to have a huge impact on national security, and if America can get AI right, we’ll be the ones to maintain a national security edge on our adversaries. But the converse is also true. We may lose that edge if we don’t get it right, or if we’re complacent and let others get ahead of us. Uh, if we get AI right, we’ll be the ones who benefit most from the economic potential, the new jobs that come from AI. But it also, and I would just say, even though this is mainly a national security and that kind of audience, we all have to be thinking about, uh, not just new industries, but also the millions of workers, many on the lower end of the economic spectrum, who could have their jobs or ways of life disrupted by AI. We’re not going to get the kind of large government support that we need, and I will talk about in a minute. Unless average Americans feel they’re being paid attention to and taken into account as we move forward um, that was a lesson of international trade 30 years ago and we can’t repeat the same mistake now. Now, also importantly, if we can lead the world in developing and implementing this new technology, we will set the norms and boundaries for the use of AI around the world. Just think of what it would mean for Russia or China, most likely, to lead the world in AI. President Putin has said that the nation that leads in AI will be the ruler of the world. And he’s made his intentions clear. China runs a massive surveillance state that relentlessly tracks its citizens, controls the flow of information, suppresses minority views, and brutally punishes dissent. So we need not to have their motivation or their way, but we need to match and exceed their level of commitment. Which, thus far, we have not done. The future where America leads the world in AI is one that recognizes privacy rights and individual liberty and democratic values. It’s a future where the militaries are smarter, not just more lethal. On the flip side, as I said, a dark cast to the future of AI, if an authoritarian uh, country wins out. President Kennedy said in his famous speech, dedicating the United States to the Space Race, “Like all technology space science has no conscience of its own. Whether it will become a force for good or evil, good or ill depends on man and only if the United States occupies the position of preeminence can we decide whether this new ocean will be a sea of peace or a terrifying fear of war.” That’s almost an exact analogy, 50 years later, to AI. So, as we await your final report, I’ve been doing some thinking about how Congress can ensure that America leads the way in AI. And I’m going to share some ideas with you. We’re going to need as much talent as possible to keep our competitive edge in AI. And that’s one reason we need some kind of comprehensive immigration reform. I was at, I won’t say which, but I was just at a re– one of the lead, the cutting edge groups doing AI. It’s a private sector group trying to make money, and there are about 220 scientists there. I said, “How many of you were born in America?” 22. Out of two hundred and some odd. Our immigration laws– restricting people who want to come here, is hurting us dramatically in this area. And we have to do something about that. In the, in Immigration Reform Bill that John McCain and I had written and that passed the Senate overwhelmingly and then was blocked in the House. Every foreigner who studied STEM and got an MA or PdD in STEM, uh, in an American university got a green card stapled to their diploma. Can you imagine how that would change the country and help us as we seek to take the lead on AI? And along similar lines, we must bolster America’s STEM education from grade school through grad school and recruit a top-notch AI-ready work force. At the moment, Data Scientists on average have a 10 years of experience in the field and China’s have five. We need to maintain and expand this advantage. But finally, and this is what I want to focus on a little bit, or have you focus on in what I’ve said today, we need to invest heavily in AI. And in other related technologies, whether its quantum computing, G5 or anything else like that. Cyber security, robotics, biotechnology. In the old days, companies could afford to do much more of the pure research. We had a Bell Labs, not too far from where I lived, I was in New York. There were monopolies, so they could afford to put billions of dollars into pure research, even if the head of– The head of whatever it wa- New- Bell, Bell, Bell! Can’t even remember what it was called back then. Um, couldn’t even see a, a way that that would create profit but we’re not, we’re not in that role anymore. Our companies don’t have that kind of ability. And so, what we need to do is have the federal government have a dramatically increased investment in AI, in the basic research, in the pure research that the companies won’t do. We’re great here in America and our companies are great at taking that kind of basic and pure research and turning it into practical things. But we’re not, we’re gonna lag behind on the basic research and the pure research. So, here is the proposal and it’s not firmed up that we would like to propose. We’re not sure the structure, but I’ll start out with the structure. What we’re thinking about it maybe having the NSF have a subsidiary “NSTF” National Science Tech Foundation, Tech, Tech Foundation. Working in concert with DARPA, which would be sort of the analog defense agency and it would be responsible for funding fundamental research related… …to AI and some other cutting edge areas like I mentioned, quantum (inaudible), quantum computing and G5 and robotics. (sniff) The NSF would also have a Board of Directors. This new agency, this combined new group would have a Board of Directors to oversee them, but the key here would be over five years, for the federal government to invent, to invest a hundred billion dollars into funding this research. And the group would fund it. Some of it would go to universities. Just like NSF does now. Some of it could go to companies. Some of it could even go to special government agencies, in the defense department or not. But the kind of pure research that needs to be done would have a long red– revenue stream. We’d also fund undergraduate scholarships, graduate fellowships, trainees, and post-doctorate support in the targeted research areas. And the development of test bed type of facilities that would be before you get to the pure research. Now, this is just a discussion draft, but it’s gotten a lot of support from the people I’ve show it to, who know this. We will do better, dollar for dollar, than the Chinese government in investing in AI. But, if they outspend us 3, 4, 5 to 1, which they’re doing now, we’ll fall behind, in five years or ten years and we will rue the day. Rue the day. Now, this idea, I should tell you, has some support of some people very close to the President, and very close to Mitch McConnell. But thus far, they have been unable to get their full-throated support. And anyone here who has any relationships with those people or people near them should be pushing this, this should be bipartisan. This should not be a partisan issue. This is about the future of America. This should not be a political issue. This should be sort of with probably even greater importance like putting a man on the Moon. But if we don’t do these real investments, in our universities, in our companies, in our government agencies, so that we become or we maintain our cutting edge lead in AI and in these other fields, it’s going to really hurt us dramatically, within several years. So, Eric has seen this proposal and, uh, has helped us modify it and if you have thoughts or ideas, please send them through the commission to us. But it’s one of my goals as minority leader, might be one of my goals as majority leader one day, (laughter) Um, to get this done because I love America, and this is so vital for the future of America. Thank you very much, everybody.