China will lead the world in artificial intelligence if the U.S. fails to spend billions of dollars more on research.
Former Google chief executive Eric Schmidt sounded the alarm bells last week, describing China’s potential leadership in A.I. as a security threat that could lead to “high-tech authoritarianism” worldwide.
“If we don’t act now, in 10 or 20 years we’ll say, ‘How could we have missed this?’” Schmidt said during an online event hosted by the Bipartisan Policy Center (BPC) think tank.
Schmidt, now the chair of the U.S. Department of Defense’s Defense Innovation Board, says that increasing government funding in A.I. research is crucial because it “can’t be made up by private philanthropy.” The amount of money needed to fund A.I. research projects—those that may not have an immediate impact—is so great that no company (not even Schmidt’s former employer) can provide a substitute.
Martijn Rasser, senior fellow for public policy at the Center for a New American Security (CNAS), an influential think tank, shares Schmidt’s view. He said that his group is urging the U.S. to spend $25 billion annually on A.I. research by 2025. For comparison, the White House said in February that it would increase non-military A.I. research spending to $2 billion annually by 2022.
“A.I. is going to be such a fundamentally enabling technology that we cannot afford to shortchange ourselves,” Rasser said.
Politicians and Schmidt, who has invested in A.I. startups and has a huge holding in Google-parent Alphabet, appear to be stoking fears about China’s dominance as one way to create a sense of urgency about increasing national A.I. funding.
Rep. Will Hurd (R-TX)—who, along with Rep. Robin Kelly (D-IL), CNAS, and the BPC, has been working on research intended to help Congress create an A.I. national strategy—cited China as the biggest threat facing the U.S. several times during the talk.
“They don’t care about civil liberties,” Hurd said about China’s hunger for online data to train its A.I. systems at the expense of people’s personal privacy. Some politicians are concerned that China’s views about A.I. and privacy will spread around the globe if the country becomes the world’s leader in A.I.
U.S. tech giants like Facebook and Google, whose executives lawmakers recently grilled about alleged anti-trust violations and weak consumer data protections, were barely mentioned during the event. These tech giants lead other U.S. companies in A.I., putting lawmakers in a predicament over whether they should break them up or impose significant fines that could limit their progress in A.I., potentially strengthening China and its own tech giants.
Hurd had one telling comment that gives a flavor of what some lawmakers are thinking regarding action against the tech giants as it relates to A.I.
“I don’t believe that breaking up some of the great American companies is the way we’re going to out-compete China,” he said.
Schmidt echoed Hurd’s comments, saying that the government needs to be sure that “American companies be the ones that dominate us globally.”
A.I. IN THE NEWS
Quantum (financial) leap. The White House said that it would invest $625 million into five new research centers that specialize in quantum computing, which researchers are hoping will be the next evolution of modern-day computing. Fortune’s Robert Hackett has the details:
The institutes, which include participation from major research universities and tech companies, will be led by Department of Energy's Argonne, Brookhaven, Fermi, Lawrence Berkeley, and Oak Ridge national laboratories. The investment is part of $1.2 billion earmarked by the National Quantum Initiative, a bipartisan bill that was signed into law last year. The remaining funds, worth $300 million, are set to be supplied by partnering corporations such as IBM, Microsoft, Intel, and others.
It’s a bird, it’s a plane… A standard Cessna 172 four-seat airplane outfitted with autonomous-flying software was able to fly on its own for 15 minutes last fall, in what was believed to be "the first complete civilian unmanned flight over a populated area,” Eye on A.I. scribe Jeremy Kahn reported. The feat was revealed by Reliable Robotics, a startup founded by former SpaceX and Tesla technologists that has exited out of “stealth” and aims to specialize in autonomous cargo deliveries. The startup’s CEO Robert Rose said that he believes the company will be able to use its technology for autonomous deliveries within two years. From Jeremy’s report:
So far, Reliable has had to have an observer on the ground directly watching its airplane fly because of FAA rules that stipulate civilian drones must be flown within direct line of vision of a human operator. But Rose believes this regulatory barrier can also be overcome to let autonomous aircraft fly far greater distances. “I believe it can all be done within the existing regulations,” he says.
China’s new A.I. rules. China is imposing new restrictions on A.I. technologies that would require “local commerce authorities” granting licenses to companies exporting their A.I. out of the country, The Wall Street Journal reported. The move could complicate the potential sale of ByteDance’s TikTok video-sharing service to U.S. acquirers like Microsoft, Walmart, and Oracle, the report said.
China last made changes to its technology export list in 2008, the Commerce Ministry said in a statement posted on the Chinese government’s website. It said adjusting the list was imperative, given the rapid development of science and technology and China’s “continuous improvement” in industrial competitiveness.
Amazon’s delivery drones get a thumbs up from the FAA. The Federal Aviation Administration has given Amazon approval to operate its Prime Air delivery drones, CNBC reported. From the article:
The approval will give Amazon broad privileges to “safely and efficiently deliver packages to customers,” the agency said. The certification comes under Part 135 of FAA regulations, which gives Amazon the ability to carry property on small drones “beyond the visual line of sight” of the operator.
Making WebEx smarter. Cisco said it would buy the startup BabbleLabs, which specializes in using machine learning to eliminate background sounds and unwanted noise from audio and video clips. The networking giant said that the BabbleLabs team would join Cisco’s WebEx video conferencing unit. Cisco did not reveal how much it spent on BabbleLabs, which raised about $18 million, according to deal-tracking service Crunchbase.
EYE ON A.I. TALENT
Scale AI has hired Brad Porter to be the computer vision software startup’s chief technology officer, Bloomberg News reported. Porter was previously the top robotics officer at Amazon, working on projects like warehouse robotics, drones, and delivery robots. Porter was one of a select few Amazon employees to be given a “distinguished engineer” title. In a LinkedIn post, Porter explained that his new role at the startup will give him time to work on the “unsexy part of AI,” referring to data training and the underlying machine learning infrastructure that powers more exciting uses.
“The sexy parts of AI are the fancy algorithms and powerful silicon that allow AI practitioners to build a model that attempts to match the performance of humans,” Porter wrote. “But if you don’t find a way to capture the performance of humans in an accurate and computer-understandable manner, it is extremely hard to build a model that matches that performance.”
Facebook hired Ankit Jain to be a senior research scientist in machine learning. Jain was previously a senior research scientist at Uber’s A.I. labs.
EYE ON A.I. RESEARCH
An A.I. bias toolkit for everyone. LinkedIn has made its LinkedIn Fairness Toolkit (LiFT) available for the public to download via GitHub. The software and methodology is based on the social networking service’s research into A.I. fairness and ethics and is intended for others to use to “detect and monitor bias” in A.I.-powered products. From LinkedIn:
The LinkedIn Fairness Toolkit (LiFT) library has broad utility for organizations who wish to conduct regular analyses of the fairness of their own models and data...
The LinkedIn Fairness Toolkit (LiFT) has already been used to measure the fairness metrics of training datasets for several models prior to their training at LinkedIn. Although we only have a limited subset of member demographic data, our use of the tool has been encouraging so far.
FORTUNE ON A.I.
50 new planets, including one as big as Neptune, are identified using A.I.—By Robert Hackett
Amazon jumps into smart wearables market with Halo activity tracking band—By Aaron Pressman
Why Palantir’s departure won’t doom Silicon Valley—By Michal Lev-Ram
Mark Zuckerberg: Facebook made an ‘operational mistake’ with Kenosha militia group—Danielle Abril
Pig brain food. Elon Musk’s brain technology startup Neuralink revealed that its human-brain computing interface can be used to monitor a live pig’s brain activity in real time. During a live demonstration, a pig named Gertrude was shown frolicking in a small enclosure while the startup’s surgically implanted embedded brain device tracked the swine’s “neural spikes.”
It was quite the sight, and it showed that Neuralink’s technology appears to work. But it’s still a small step in the company’s quest to make technology that could help restore brain functions with people with neurological disorders.
As Fortune’s David Morris reported, some researchers believe that “the company isn’t offering much that’s new, and that its biggest speculative claims aren’t grounded in current science.” Andrew Jackson, a professor of neural interfaces at Newcastle University, questioned the startup’s claims that the technology could eventually help record memories.
From the report:
“There is a big difference between recording brain cells and ‘reading thoughts,’ especially when it comes to higher-level cognitive functions we don’t understand as well,” Jackson said. “The idea of ‘writing to the brain’ is even more questionable…There are fundamental limitations to targeting specific networks of neurons in a meaningful way using electrical stimulation.”
And before you go, I'd like to introduce you to our Fortune video series, AI Minute. Check out this video where Jeremy talks about the relationship between A.I. and poetry.