Artificial Intelligence: The Global Race for the New Frontier – Narrated by David Strathairn

Artificial Intelligence: The Global Race for the New Frontier – Narrated by David Strathairn


-: I’m concerned with the world in which we’re going to live tomorrow. A world in which a new machine, the digital computer, may be of even greater importance than the atomic bomb. Exciting and challenging–but doesn’t it worry you? Narrator: New technologies are transforming our world. An automation revolution is making daily life easier for millions of consumers. -: Hello, Siri, what’s the weather going to be today? -: Alexa, what can I do about my migraine? -: Hey, Google, open YouTube on the living room TV. -: Alexa, turn on bedroom light. It’s that simple. Narrator: But it also fueling international rivalries, displacing workers from farms and factories, and expanding the reach of authoritarian regimes. Great Decisions asks what the international community can do now to prepare for the looming challenges of the coming digital age. Robotic voice: Artificial Intelligence: The Global Race for the New Frontier, next on Great Decisions. (dramatic music) Announcer: Great Decisions is produced by the Foreign Policy Association, in association with Thomson Reuters. Funding for Great Decisions is provided by PricewaterhouseCoopers, LLP and the Nelson B. Delavan Foundation. (electronic music) Narrator: By the end of this broadcast, internet users around the world will have sent nine million tweets, posted over a million photos to Instagram, and searched Google 102 million times. In their homes, they will have used smart speakers, smart fridges, smart thermostats, and even smart beds. In all, they will have generated a staggering 46 quadrillion bites of new data. -: Every day billions of dollars change hands and countless decisions are made on the basis of our likes and dislikes. Narrator: All this data makes it possible for computer scientists to create programs that are capable of “learning” on their own. Those programs constitute artificial intelligence, or A.I. -: Computers mostly learn from access to large amounts of data. Data is the fuel for artificial intelligence, just like gasoline is the fuel for a car. The more gasoline, the farther the car can travel. The more data, the faster a computer can get to its conclusion of reasoning successfully. -: I will generally define A.I. as technologies that rely on data, they’re designed to find patterns in data, and, based on an understanding of those patterns, to make predictions about the world. -: Data is the lifeblood of these types of algorithms. You have to have a sufficient volume of training data to make them do the things that they do. Now, what’s happened is, over the last three decades, we’ve begun digitizing information on a scale never seen before. Narrator: Today, applications of A.I. are everywhere. -: Voice recognition, machine translation, image recognition, diagnoses of certain diseases. Researchers are working on autonomous vehicles. Robots that don’t need to be explicitly programmed to perform certain tasks. -: If you’re watching a show on Netflix, you’re going to see recommendations for what you might want to see next. That’s based on machine learning. If you’re driving an automobile that has a camera that may have a sensor when it identifies that there’s a pedestrian crossing, that’s artificial intelligence. -: There’s an enormous amount of A.I. going into what information you’re being served up on your Twitter feed or your Facebook news feed. It’s having a huge role in what’s advertised to you. Already we’re starting to live in this “A.I. world.” Newscaster: We turn now to Beijing, where a computer program developed by a subsidiary of Google has defeated the world’s top-ranked player at the board game go. Newscaster: A spokesman for Facebook promised today that the platform will not take down a deepfake video of Mark Zuckerberg, which had been created by two artists in conjunction with an Israeli artificial intelligence startup. -: Spectre showed me that whoever controls the data controls the future. Newscaster: On the lighter side tonight, we bring you a story of “fishal recognition.” Reporter: Scientists in Washington state have trained a computer to identify different species of salmon as they pass over Chief Joseph Dam en route to upstream spawning grounds of the Columbia River. Narrator: As the use of A.I. expands, workers are anxious about the prospect that new advances in automation could make many jobs obsolete. -: You’re already seeing automation sweeping the workforce, but if you look closely it’s not really eliminating jobs. What it’s often doing is diminishing the value of work. -: A lot of the developing countries depend on manufacturing. They produce our shoes, our t-shirts, where you have a lot of low-skilled laborers doing these things. However, if robots can be used in manufacturing, then these developing countries, these workers will lose out. And some countries might never industrialize at all. (gentle music) Narrator: Two countries stand at the forefront of the A.I. revolution: the United States and China. -: I just got back from Beijing, and what I saw was an industry that was pretty much on par with the U.S. What singles China and the U.S. out is that they can actually support these massive tech companies that are able to accrue the large-scale computational infrastructure and the massive amounts of data that is needed to create A.I. at scale. -: The Chinese government has articulated in formal documents, in white paper strategies that they’ve been kind enough to publish, that it is their intent to supplant U.S. dominance in ten key technologically-driven industries, including artificial intelligence. Narrator: With relations between Washington and Beijing growing tense, American experts have presented A.I. as a critical new battlefield. -: We are engaged in an epic race for A.I. supremacy. -: Big investments in A.I. by China and Europe already threaten U.S. dominance in this field. -: Both our allies and potential adversaries are pursuing A.I. dominance. -: By some accounts China is investing $7 billion in A.I. through 2030. -: Sometimes technological races end up being wars between civilizations. And so, if we think now that there’s really a contest between the Western democratic style and the Chinese, the consequences of the A.I. race could be consequences for the future of humanity. -: Who’s gonna develop and lead the way in these technologies and, importantly, who’s also gonna try to shape rules around how they’re used? And right now we’re not doing either. We’re not at the cutting edge anymore. I think the Chinese are at least where we are. And we’re certainly not trying to establish the rules under which these technologies are to be used. Narrator: Some observers worry whether private American corporations can keep pace with Chinese companies that are backed by massive government spending on research and development. -: If the United States doesn’t figure out that we have to have an answer for the Huawei model, then China’s going to do the same thing to us on A.I. that they did to us on 5G. That means the United States has to get over its reticence to have a true industrial policy where the government and the private sector work together to try to develop new technologies. -: We’ve been like, well, we got Google, we got Microsoft, we got Amazon, they’ll fight this race. They have a lot of great engineers, but they have their limits. They’re budgets are big, but they’re not like a nation state budget. -: We have this myth in the tech sector, it’s the proverbial person, couple of people, they’re inventing in a garage or a dorm room and then they go on and their companies become Microsoft or Google or Hewlett Packard. There is an element of truth to that, but it misses something of fundamental importance. Since World War II, the federal government has funded basic scientific and technological research, typically through the large research universities in the nation. We’re going to need this for A.I. So, when some people say we need more federal funding for A.I., I think they’re right. -: We’ve been leaders in innovation traditionally because we’ve made investments in our education system, we’ve supported our universities and researchers in creating breakthrough technologies, and we need to make those investments if we want to continue to be leaders in innovation going forward. Narrator: Other experts question whether framing A.I. as a competition between China and the U.S. might have dangerous consequences. -: I’ve seen this narrative around China used a lot of times to kind of derail that discussion. Like, we can’t talk about ethics, can’t talk about guardrails, can’t talk about regulation because if we do implement those things, China’s going to race ahead toward whatever this goal is and we’re going to lose. Maybe we don’t want to be racing toward shoddy, exploitative, human rights violating, ubiquitous technologies. Maybe that’s not a race we want to play in. Maybe we want to do something different. -: I’m very concerned about the tenor of the discussion surrounding technological innovation and how it’s some kind of new realm of security competition between the U.S. and China. This is already happening without much discussion or regard to the implications for both the global economy and the future of innovation. (soft music) Narrator: Within China, technologies based on A.I. have made possible some alarming new methods of surveillance and social control. Newscaster: We being our coverage tonight in China’s northwestern region of Xinjiang, where a data leak has revealed that the government is regularly tracking the whereabouts of some 2.6 million citizens. Newscaster: Sources tell us that one social credit system now being tested in China scores citizens’ reputations on a scale of 350 to 950 based on what they buy, where they go, and how they behave. Newscaster: Breaking news tonight out of Hong Kong, where crowds are rallying in support of a student who was arrested for carrying ten laser pointers, which have been used by protestors to sabotage security cameras and thwart police surveillance. -: There’s actually a recent study done that found that eight out of the top ten most surveilled cities in the world are in China. And when you think of a country with a population size of China, you can imagine how many cameras that actually is. -: A.I. certainly has enabled surveillance. Instead of having humans watch hours and hours of video from CCTV cameras, you can use an algorithm to identify faces, to track movements. -: We don’t really know the future trajectory of autocracy in the 21st century. What we do know is that artificial intelligence and other technologies might in fact enable autocracy rather than challenge it. (soft music) Narrator: When China exports new surveillance technologies to the developing world, those technologies might help fight crime and promote public safety, or they might enhance the power of authoritarian regimes. -: China is interested in selling high-tech products. They’re less concerned about how they’re used. And they’re willing to market them to whatever government happens to be in place in any given country. -: Open Technology Fund found that Chinese internet technology for surveillance or censorship had been exported to 73 countries around the world across five continents. And it’s important to keep in mind these are not only the autocrats. It is also in some cases democracies, but a lot of countries that are “partly free,” and these technologies can really tip the balance. -: Some governments, for example in Uruguay, they’re having all sorts of trouble at the moment with security. Crime is on the rise, and Uruguay really needs some surveillance systems, some CCTV and other systems, to monitor things and ensure a degree of stability on the streets, and will likely use these technologies precisely for that purpose. But in other countries, those that happen to have more authoritarian tendencies, China’s marketing these technologies differently. It’s marketing them in a way that is promoting of population control. (spooky music) Narrator: Some commentators argue that fears over Chinese technology exports are overblown. -: The notion that is being spread is that surveillance systems are a Chinese invention and a Chinese monopoly. Well, that is not the case. Surveillance systems are all over the world. -: These are very mundane technologies: cameras, digital facial recognition technology, other kinds of internet filters and VPN systems, etc., and they’re not controlled by governments. They’re sold by commercial entities. -: On the one hand, these are dual-use technologies, but on the other hand, they are being configured already in China in a politically repressive way, and that means that, when they’re exported, they’re more likely to be used in that way. Narrator: By pouring resources into A.I. research, Chinese authorities have created an almost insatiable appetite for data. -: Cloudwalk is a Chinese A.I. firm that does facial recognition. They had only tested it in China. The Zimbabwean government bought it– got it for a lovely discount– and has used it for all sorts of repressive behavior. -: And what Zimbabwe offered to China was basically its data. It offered faces–lots and lots of faces–that could be used to refine China’s facial recognition algorithm. -: The data-driven economy, and in particular A.I., is going to require really good governance, and a lot of developing countries lack the funding, the willingness, and the ability to regulate this effectively. (upbeat music) Narrator: The data arms race with China is driving calls for more open sharing of information. -: As a nation, we in the United States are going to have to compete with China even though we only have about a quarter of China’s population. So, over time, we should expect China to produce more data than the United States. We need a national strategy. We need a foreign policy strategy that will allow the U.S. to connect with other like-minded nations. -: More and more people have thought about building up walls here. Not just China. There are other countries around the world, too, who have wanted to build up walls around their networks. That basic architecture of the internet, in terms of the free flow of data, has been very, very important and we need to preserve that. Narrator: The modern economy relies on the free flow of data around the world. But the global trade rules governing that flow were adopted when the internet was in its infancy. -: The most comprehensive trade agreement in the world is the World Trade Organization. It contains trade agreements governing services such as telecommunications, banking that rely on cross-border data flows, but the WTO never mentions data. So, they talk about “computer services,” but they don’t talk about the internet because they were written before the internet became widely used. -: There is the General Agreement on Trade in Services. It hasn’t really been enforced by the WTO or by parties to the WTO to govern data flows. But I think if you really took it seriously, it would, for example, say that China’s blocking of Facebook or Wikipedia, parts of Google, is completely illegal under international trading rules. So, there’s a lot of potential for that to become a major thing. It just hasn’t yet. (electronic music) Narrator: Recognizing the importance of data-driven technologies, both the Obama and Trump administrations have tried to include provisions in new trade deals that would keep the internet unrestricted. -: Once approved by Congress, this new deal will be the most modern, up-to-date, and balanced trade agreement in the history of our country. Likewise, it will be the most advanced trade deal in the world with ambitious provisions on the digital economy. -: Digital trade is really a new issue. The United States participated in writing a digital chapter for the first time in the Trans-Pacific Partnership, but we are not a signatory to that. There’s work happening again to have a digital chapter in the new version of NAFTA, USMCA. That agreement would be the first. -: I don’t know that we’ll necessarily need to have this addressed in every trade agreement. I think what we need to take more care to do is ensure that we’re not pursuing domestic laws or policies that block the flow of data. (gentle music) Narrator: But the free flow of data also brings concerns about privacy. So far, the European Union has led the way in formulating regulations to protect users’ privacy, particularly through the General Data Protection Regulation, or GDPR, of 2018. -: We should celebrate the transformative work of the European institutions tasked with the successful implementation of the GDPR. It is time for the rest of the world, including my home country, to follow your lead. -: I think that the first thing that we need to do in the United States is make sure that we’re protecting consumer privacy and then make sure that we’re protecting the rights of consumers online. And if we’re going to have a seat at the table internationally, we need to have domestic laws in place and then make sure that we are exporting our views. -: There is a challenge as we think about artificial intelligence and the protection of privacy: namely, the tighter the privacy protection, the harder it may become to access large data sets for renewed research in A.I. So, how do we strike the balance? That’s what people in Europe are starting to talk about. I think it’s going to require a new wave of privacy protection in the 2020s. -: It is true, I think, if you have strong privacy laws, you will restrict A.I. from becoming good at certain things. Maybe if privacy laws prevent us collecting so much information about buying decisions, so much geared towards advertising, then maybe A.I. will start to develop approaches to other problems. And maybe it would move A.I. efforts towards other goals having less to do with human manipulation and behavioral control. (upbeat music) Narrator: In the coming decade, Congress and the White House will enact laws that will help determine how A.I. develops in the U.S. and around the world. But experts are concerned that neither voters nor legislators have the sophisticated understanding of the challenges of A.I. necessary to make informed decisions. -: How do you sustain a business model in which users don’t pay for your service? -: Senator, we run ads. -: I see. -: Well, if we’re going to legislate on technology, then we need to get members of Congress up to speed on these issues. My background is in technology, so I consider myself one of those folks who can help educate my colleagues. -: How does search work? -: Is Twitter the same as what you do? -: And next time you visit, if you would please bring some fiber because we don’t have connectivity. -: So, it’s not a little man sitting behind the curtain figuring out what we’re going to show? -: The problem with Congress isn’t that they’re idiots. It’s that they don’t do anything. Congress is about as relevant as, I don’t know, the local deli to this conversation. They talk a lot more. They occasionally will write some bills, but last time I checked, I haven’t seen Congress pass any bill of any relevance to this. -: I have a seven-year-old granddaughter who picked up her phone before the election and up on there pops a picture of her grandfather. And I’m not gonna say into the record what kind of language was used around that picture of her grandfather, but I’d ask you how does that show up on a seven year old’s iPhone who’s playing a kid’s game? -: iPhone is made by a different company and so… -: It might have been an Android, it’s just it was a hand-me-down of some kind. -: The sort of framing of why are these senators so behind the times is a little annoying to me. We’ve lived through like two decades of mystification and obfuscation by the tech industry. The tech industry claims to own the future and anyone who dared to suggest that we might want to regulate it, that we might want to bend that future to norms that should be controlled by democratically elected governments, was considered way out of line. How dare you hamper innovation? It’s no excuse that people who are in charge don’t do a little bit more reading, but I think this is something that does need to land back on tech’s doorstep. Narrator: Experts and policymakers are deeply divided over how much to regulate companies working to develop A.I. technologies. -: I do think that the government should play a role in regulating, certainly, the deployment of A.I. applications. It’s not just enough for tech companies to regulate themselves. -: Most people would prefer that the private sector and the public sector have an informal agreement and work collaboratively without overburdensome government regulation because the reality is that government regulation is not agile and it often constrains the type of innovation that’s going to be necessary for national security going forward. -: I would be more concerned about having a discussion with U.S. companies about where are you selling this artificial intelligence? That’s the kind of discussion I think should be taking place with our tech companies, and my suspicion is it’s not happening. Narrator: Multinational institutions, such as the Organization for Economic Cooperation and Development, have put forth principles for the responsible use of A.I. But there is concern that no tools exist to enforce these axioms. -: All firms in those countries, when they utilize A.I., when they utilize data, should follow ethical business practices that are agreed upon by many countries. And that’s a good thing, and principles often lead to rules. -: I don’t have optimism that they will be implemented simply because we said they should be. I think it’s going to take building different forms of power to sort of force the implementation of these principles. It will take things like the waiving of trade secrecy for some of these critical systems so that we can examine them and validate them. -: Fundamentally, artificial intelligence empowers machines, computers, to make decisions that previously have only been made by human beings. When we think about it in those terms, I think we quickly recognize that one of the fundamental questions for all of us is what kind of decisions do we want machines to make? If in fact, we want the best of humanity to be reflected in these decisions made by machines, we need the companies that design this technology to be guided by clear ethical principles, rules, and oversight. Narrator: Because new technologies can so easily cross national borders, it will be imperative for countries to cooperate to meet the challenges those technologies create. World leaders once came together to construct a framework to manage the proliferation of nuclear technology. Time will tell whether they can do so again as the world enters the year of artificial intelligence. (computer sound effects) (upbeat music) -: Great Decisions is America’s largest discussion program on global affairs. Discussion groups meet in community centers, libraries, places of worship, and homes across the country to discuss global issues with their community. Participants read the eight-topic briefing book, meet to discuss each topic, and complete a ballot, which shares their views with Congress. To start or join a discussion group in your community visit fpa.org, or call 1-800-477-5836. Great Decisions is produced by the Foreign Policy Association, in association with Thomson Reuters. Funding for Great Decisions is provided by PricewaterhouseCoopers, LLP and the Nelson B. Delavan Foundation. (dramatic music) (upbeat music)