We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 - 28. Join AI and data leaders for insightful talks and exciting networking opportunities. Register today!
AI has the potential to improve human lives and a company’s bottom line, but it can also accelerate inequality and eliminate jobs during the worst U.S. recession since the Great Depression. This dual promise and peril led members of the House Budget Committee to hold a hearing today to discuss the impact of AI on economic recovery, the future of work, and the federal budget.
Expert witnesses recommended approaches that ranged from giving people lifelong upskilling accounts to creating regional investment districts and portable benefits.
MIT professor and economist Daron Acemoglu warned the committee about the dangers of excessive automation. Acemoglu recently found that every robot replaces 3.3 human jobs in the U.S. In a working paper published by the National Bureau of Economic Research, Acemoglu detailed how excessive automation looks for ways to replace workers with machines or algorithms but produces few new jobs. Companies are currently incentivized by a U.S. tax code that taxes capital at a lower rate than human labor, policy he said incentivizes companies to replace humans with automation. In practice, this can be as simple as replacing a McDonald’s worker with a touchscreen. He argues automation has been a drag on the U.S. economy, potentially slowed market productivity, and failed to lead to higher wages for low- and middle-class workers.
“AI is a broad technological platform with great promise. It can be used for helping human productivity and creating new human tasks, but it could exacerbate the same trends if we use it just for automation,” he said. “Excessive automation is not an inexorable development. It is a result of choices, and we can make different choices.”
Also shaping automation in the economy are Big Tech companies with business models that focus on replacing human labor with algorithms, and priorities enshrined in U.S. government funding. And Acemoglu said Congress must consider factors unseen in data, such as the loss of self-worth a person suffers without a job or how employment impacts the likelihood of a person becoming involved in their community.
Rep. Bill Flores (R-TX) cautioned against changes to how the government treats capital and labor, saying “I think we as policymakers need to be very careful about trying to get into adjusting the mix of capital versus labor because as you said early on Chairman [John] Yarmuth, the government moves slowly. And I think we as policymakers could end up being well behind where the economy is if we’re not careful with that.”
It’s unclear whether testimony delivered today could influence any particular upcoming piece of legislation, but it could impact how members of Congress shape policy related to funding in areas like research and development, national defense, and job retraining.
The McKinsey Global Institute predicts AI will drive GDP growth around the world in the decades ahead, but studies also find that automation in the workplace is expected to particularly impact Black or Latinx people, who make up a larger percentage of the service economy than other groups. Certain cities, states, and regions are also expected to be disproportionately impacted by the proliferation of AI. About one in three jobs in the U.S. may require some form of retraining due to automation, according to the OECD. On the business side, a KPMG report released last year urged company executives to take AI-driven job loss seriously and to consider ways to retrain or reskill workers.
High-speed broadband access received bipartisan support in the hearing as a way to close the digital divide and ensure widespread access to education, job training, and remote job opportunities. A bill with nearly $80 billion in broadband funding was approved by the House of Representatives in July, but the Senate has not yet taken up the legislation.
Helping workers find job or skill training will be an important part of addressing the instability and job loss brought on by AI, said Stanford Human-Centered Artificial Intelligence (HAI) associate director Susan Athey, but the United States hasn’t historically done a great job of helping displaced workers impacted by automation. Athey is currently working with the Rhode Island state government on ways AI can help people find work. He said people need to feel comfortable with upskilling recommendations before they will complete job training programs and acquire new skills.
“They need to have confidence that if they do make that effort and take that scarce time and money, they will be able to use that to get a new job,” Athey said. “I’m optimistic about the future, but we have to be intentional about it, and we actually have to execute and follow through on promises.”
Stanford HAI is only a little over a year old, but the organization led by former Google Cloud chief scientist of AI and ImageNet creator Dr. Fei-Fei Li has made multiple policy proposals to Congress. Last fall, Stanford HAI urged the federal government to invest $120 billion in the U.S. AI ecosystem over the next decade, with $2 billion in funding for entrepreneurs and innovation, $7 billion for AI research, and $3 billion for education. In June, Stanford HAI joined companies like AWS, Google, and Microsoft in urging the federal government to create a national AI research cloud bill.
As has become standard in congressional hearings about artificial intelligence, competition between China and the U.S. came up on multiple occasions. Dr. Jason Matheny is director of the Center for Security and Emerging Technology (CSET) at Georgetown University and serves on the National Security Commission on Artificial Intelligence (NSCAI), a temporary body formed to provide Congress with military and national security policy recommendations. The group asserts that AI dominance by the United States is vital to national security.
Matheny said immigration of highly skilled workers was key to the U.S. winning World War II and the Cold War and that the U.S. maintains an asymmetrical advantage over China today because more AI research talent wants to live and work in the U.S. National Science Foundation (NSF) data shows more than half of Masters and Ph.D. level computer scientists in the U.S. today were born abroad. He also named immigration as an important aspect of addressing worker shortages brought on by Baby Boomers leaving the workforce.
Matheny named a U.S. lead in the semiconductor hardware used to accelerate AI as another key advantage over China.
“It will be very difficult for China to match us if we play our cards right,” Matheny said. “We shouldn’t rest on our laurels, but if we pursue policies that strengthen our semiconductor industry while also placing the appropriate controls on the manufacturing equipment that China doesn’t have and that China currently doesn’t have the ability to produce itself and is probably a decade away from being able to produce itself, we’ll be in a very strong position.”
He cautioned that the U.S. is still too reliant on semiconductor hardware from abroad and called Intel potentially working with Taiwan Semiconductor Manufacturing Company to manufacture advanced semiconductors outside the United States “extremely worrying.” As part of securing the semiconductor supply chain, Matheny also encouraged strengthening alliances with other nations. In June, the U.S. and more than a dozen other countries joined the Global Partnership on AI (GPAI). The group holds its first annual meeting in December.
The NSCAI first recommended that Congress invest in public-private partnerships to bring more semiconductor hardware back to the U.S. earlier this year. On the subject of AI talent and retraining, the NSCAI earlier this summer recommended creating a government university to train people for AI work in the federal government. The NSCAI will issue a final list of recommendations to Congress in spring 2021.
For markedly different reasons, members of Congress voiced some criticisms at the hearing today. In her questioning, Rep. Sheila Jackson Lee (D-TX) said “We’re all speaking to ‘haves’ because the ‘have nots’ are not in the room.” This was in reference to the fact that the witness list did not include the voices of people impacted by job dislocation or inequality. Last fall, for example, the Federal Trade Commission received a complaint about AI-powered hiring tool HireVue that alleged unfair practices and intrusive collection of biometric data.
In response to a question from Jackson Lee, Brookings Institution director of governance studies Darrell West said investment in broadband, telemedicine, and education could help address inequities in tech.
“We’re almost in a situation where technology is helping to fuel the inequality in the sense that the haves are doing better and getting tax breaks and have programs that support them, and people at the lower end aren’t even in the game. They don’t have access to the digital economy,” he said.
In his critique, ranking member Steve Womack (R-AK) called the hearing focused on AI interesting and important but said the budget committee would be better served by focusing on how to reduce the U.S. federal budget deficit. Updated projections released by the Congressional Budget Office Wednesday found that the deficit now stands at $3.3 trillion, more than triple what it was in fiscal year 2019 and the highest percentage of debt compared to GDP since 1945.
When asked by Womack how the federal government should prioritize research and development funding, Matheny suggested a focus on areas where private businesses may be likely to underinvest. This includes basic research carried out by groups like DARPA and the National Science Foundation and testing and evaluation by agencies like the National Institute of Standards and Technology (NIST).
Numerous witnesses testifying before the committee today said the government can use AI to improve government efficiency.
“I do believe it is a good time to start thinking about modernizing federal government infrastructure,” Athey said, adding that cybersecurity, fraud protection, and administrative services are ways the federal government can deliver services. A joint New York University-Stanford study examining federal government use of AI found that only 15% of federal agencies currently use advanced forms of artificial intelligence today. Some government use of AI might result in job loss, but Athey argued job loss could lead to better allocation of human resources for society.
“While [AI] loses a piece of employment for the worker sitting behind the counter, maybe there’s other things your government could be doing: more childcare, more eldercare. There are other services that are underprovided, where the human workers can be better deployed if we use technology to do things where the human is kind of getting wasted on both sides of the table,” she said.
In other recent government regulation and AI news, the Portland City Council passed the strictest facial recognition ban in the United States Wednesday. Businesses and government agencies that fail to comply can face fines of $1,000 per day for each violation of the law.
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn more about membership.