The U.S. Leads in Artificial Intelligence, but for How Long?
Even as the world’s top artificial-intelligence researchers gathered in Los Angeles this week, many are beginning to wonder just how much longer the U.S. will remain the epicenter of AI.
The Neural Information Processing System (NIPS) conference in Long Beach is the number one place for presenting breakthroughs in AI. But U.S. government policies threaten to put a dampener on the recent boom in the field.
The U.S. Congress’s tax plan is the latest challenge, threatening to raise costs for graduate students significantly. This follows reduced funding for fields including AI and tightening of rules on immigration for international researchers.
“Through decades of public and private investment, the U.S. currently leads in AI basic research,” says Andrew Ng, a leading figure in the field who held an academic post at Stanford before leading research efforts at Google in the U.S. and Baidu in China. “Ill-advised policies can quickly squander this lead.”
The implications of the tax plan in particular will be hot topics at NIPS, an event where companies and schools woo top talent. For many years, the event was a small gathering of researchers toiling away on neural networks, a subfield of AI that had seen limited success and had largely fallen out of fashion. Around 2012, however, dramatic progress breathed new life into that area.
Reflecting the wider AI boom, over the past few years the NIPS conference has grown from a small gathering of a few hundred academics to a sprawling event with thousands of attendees, big-name corporate recruiters, and lavish parties.
The tax bill, which is currently working its way through Congress, proposes requiring graduate students to pay taxes on their tuition, which is normally waived by academic institutions. For many, this would suddenly mean a tax bill of around $10,000, making it much harder for U.S. universities and professors to attract graduate students. The bill passed by the Senate on Saturday did not include this provision, but it could still find its way into the final version.
“The tax plan is preposterous,” says Zachary Lipton, an AI researcher who will join Carnegie Mellon University as an assistant professor later this month after a stint at Amazon. “It’s a grave threat to both our competitiveness with foreign universities and academia’s competitiveness with industry.”
The U.S. government’s moves come at an inopportune moment given how competitive AI is becoming, and how much emphasis other nations are placing on the technology. Over the long term, the consequences could also be felt not just in academia but in the U.S. technology scene, which has often fed off academic advances.
Meanwhile, other governments spy an opportunity, and an imperative to invest heavily in AI research. The Chinese government, for instance, has announced plans to pour billions of dollars into research and development (see “China’s AI Awakening”).
Those who study the implications of AI are just as worried about the government’s plans. “I’m with everyone else—this is devastating,” says Erik Brynjolfsson, a professor at MIT’s Sloan School of Management who is studying the impact of AI on economic growth and inequality. “It’s honestly like it was designed by America’s enemies who want to take us down a notch. American policy is contributing to having AI research leave the country, literally.”
The tax plan also comes as big companies pour huge sums into recruiting talent. Industry giants like Google, Facebook, and Amazon routinely offer would-be graduate students $100,000 or $200,000 to join their ranks. Olga Russakovsky, an expert in machine vision and learning who recently moved to Princeton University as an assistant professor, worries that the allure of industrial jobs could skew the focus of AI research.
“Right now companies are already putting way more money into AI R&D than the government, which is already a cause for concern,” she says. “If the disparity grows even further, that’ll affect the type of research we can do, and our objectives and values.”
The government’s immigration policy is affecting AI startups, too. Karl Iagnemma, CTO of nuTonomy, an autonomous-car startup spun out of MIT, says he has had difficulty recruiting researchers in the U.S., and has focused on luring people to the company’s offices in Singapore as a result.
But the government actions are likely to hit the academic world hardest. “Students are sacrificing a lot to continue their education, to give back to society in more meaningful ways, to push the frontiers of science, and to potentially become educators of the next generation of students,” says Russakovsky. “To make their lives any harder, or deny them the opportunity to do it at all, is just terrible and a really bad idea for society at large.”
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
Deep learning pioneer Geoffrey Hinton has quit Google
Hinton will be speaking at EmTech Digital on Wednesday.
The future of generative AI is niche, not generalized
ChatGPT has sparked speculation about artificial general intelligence. But the next real phase of AI will be in specific domains and contexts.
Video: Geoffrey Hinton talks about the “existential threat” of AI
Watch Hinton speak with Will Douglas Heaven, MIT Technology Review’s senior editor for AI, at EmTech Digital.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.