Rapid Responses to AI

AI could displace middle-skill workers and widen the wealth gap, says landmark study, which calls for better data and more investment in continuing education to help workers make career pivots. Also, Colorado invests in a new language of learning, and Trump’s surprising labor secretary pick.

Ensuring That AI Helps Workers

Artificial intelligence has emerged as a general purpose technology with sweeping implications for the workforce and education. While it’s impossible to precisely predict the scope and timing of looming changes to the labor market, the U.S. should build its capacity to rapidly detect and respond to AI developments.

That’s the big-ticket framing of a broad new report from the National Academies of Sciences, Engineering, and Medicine. Congress requested the study, tapping an all-star committee of experts to assess the current and future impact of AI on the workforce.

“In contemplating what the future holds, one must approach predictions with humility,” the study says, noting that while the rapid adoption of large language models has outpaced expectations, the road to fully autonomous vehicles has proved lengthier than anticipated.

The committee’s approach is cautious, and the report has a measured, often passive tone. It also throws shade on “highly speculative forecasts” by consulting firms about AI’s effects on labor demand, and credulous reporting by media outlets on those projections.

The uncertainty described in the report echoes what we’ve heard from top experts about how jobs could change, and what that might mean for education and job training programs.

Even so, significant advances in AI technology are highly likely, the study says, with major impacts on worker productivity in the next decade.

“We should anticipate productivity gains across a wide cross section of the economy, and we are already seeing those gains in certain areas such as software development,” Tom Mitchell, a professor at Carnegie Mellon University and the committee’s co-chair, said in a statement.

Those gains will play out differently across job roles and could worsen severe income and wealth inequality in the U.S.

Beyond dramatic productivity boosts for software developers, the work of paralegals, customer service agents, and workers who summarize documents already is being reshaped by AI, the report says. Entertainment, finance, healthcare, education, retail, manufacturing, transportation, and many other industries are poised for transformation.

One plausible scenario is that AI could accelerate the information era’s polarized labor market, which has reduced demand for middle-skill office and production jobs while increasing returns for four-year degrees and, especially, advanced degrees.

“AI could accelerate occupational polarization,” the committee said, “by automating more nonroutine tasks and increasing the demand for elite expertise while displacing middle-skill workers.”

What’s clear, according to the study, is an urgent need for better data collection on AI’s use by businesses and workers. It calls for public-private data partnerships to “widely share real-time data on skills supply and demand, wages, and continuing education opportunities.”

While the committee could have taken a more active position with its recommendations, the report strongly asserts that the AI future is not preordained, including on questions about whether the technology will replace workers—or help them.

“AI is a powerful tool, but it is still a tool to be directed by humans,” said Erik Brynjolfsson, director of the Stanford Digital Economy Lab and the committee’s co-chair.

The report is a comprehensive overview, says Tiffany Hsieh, director of innovation programs with JFF’s Center for Artificial Intelligence and the Future of Work. She praised the committee’s call to be more adaptive to real-time changes in the job market, and the focus on lifelong learning.

The Kicker: “The education and workforce ecosystem has a responsibility to be intentional with how we value humans in an AI-powered world and design jobs and systems around that,” says Hsieh.

Lifetime Learning and Career Pivots

As the pace of change picks up in coming years, the study says many workers will need “just-in-time training” to improve their job prospects. Likewise, a new generation of students may require different sets of skills and training.

“Providing workers with retraining opportunities and keeping them informed of shifting demands for different skills will be essential to assuring that the workforce benefits from advancing AI technology, and that the new wealth generated from these advances is shared widely,” said Mitchell of Carnegie Mellon.

The report calls for research on effective approaches to continuing education, particularly short-term programs that teach specific skills that are in high and growing demand. It also suggests that a combination of vocational training and a liberal arts education could benefit a broad group of students in the AI era.

Sectoral employment training programs, community colleges, and registered apprenticeships get shout-outs from the committee, as do Northeastern University’s co-op program and the P-TECH K-14 model.

“Community college partnerships with consortia of local employers to provide training for emerging jobs with upward mobility potential also remain a crucial part of the postsecondary education system, both for new high school graduates and for returning students attempting to augment or change careers,” the report says.

Citing a growing body of evidence, the committee says training programs that help working learners break into skilled occupations have several key ingredients: upfront screening, occupational and soft skills training, and wraparound services.

To help Americans make anticipated career pivots, the U.S. will have to close the gap on what it spends on unemployment insurance and job training programs—just 0.3% of GDP compared to the best-in-class 3% spent by Denmark.

“New educational technology, no matter how spectacular, will not itself be sufficient to support the training needs of present and future workers, particularly displaced adult workers,” the report concludes. “Substantial public investments will be required.”

Jobs as Bundles of Tasks

One way AI will impact jobs is through its impacts on individual tasks. But widespread adoption of the technology also could lead to fundamental changes in the structure of jobs and industries, the report says. Even if longer-term consequences are favorable, the transition is likely to be economically and perhaps societally destabilizing.

The most relevant concern for workers is not whether AI will eliminate jobs, according to the committee, but how it will shape the labor market value of expertise. Will it augment the value of the skills and expertise of workers or instead erode that value by providing cheaper machine substitutes?

In some cases, the report predicts that AI will broaden the reach of workers rather than making their expertise superfluous. 

The technology will help store managers track inventory and staffing while allowing them to focus more on coaching workers, engaging with customers, and solving problems, according to an example from the study. Likewise, AI can quickly write code, but software developers will need to create and review the software’s architecture.

Mike Turner is global VP of offerings for the software and platform engineering service line at Cognizant, a major technology consulting and services firm. He has a front-row seat to the change happening in software development and engineering roles across industries.

AI is enabling—and requiring—more software engineers to work across functional domains, says Turner. Engineers have typically worked in distinct lanes, like front end, back end, and cloud development, and even on separate teams. Now they’re converging into broader full-stack development roles, and junior team members are able to make that leap more quickly.

“With generative AI, you have this new level of support and coaching. If you’re a software engineer and you’re used to building in a certain part of the tech stack, it can help you understand another domain,” he says. “There’s a lot more fluidity happening within those roles.” 
—Elyse Ashburn contributed reporting for this article.

Open Tabs

Clean Energy
Many across workforce education are nervously watching to see if Republicans will make good on threats to claw back funds under the Inflation Reduction Act, which will provide at least $390B in tax breaks and subsidies for clean energy manufacturing and EV battery production. Yet the climate law overwhelmingly benefits GOP districts, and The New York Times reports that some Republicans are seeking to preserve aspects of the Biden administration’s signature legislation.

Segregation in Tech
Computer, engineering, and math jobs remain stubbornly segregated by race, ethnicity, and gender—despite efforts to help more women and Black and Latino workers break into the tech industry, according to recent research by the Brookings Institution. Many companies are now pulling back on DEI initiatives amid cost-cutting and pressure from conservatives. That’s a problem for attempts to widen economic opportunity, the researchers write, as those “highly digital jobs” pay well.

Education Pipeline
Enrollment of 18-year-old, first-time college students fell precipitously at four-year institutions this fall, while community colleges saw smaller declines. The combined drop was 5%, according to an analysis by the National Student Clearinghouse Research Center. Regional demographic and labor market trends, the Supreme Court’s ban on affirmative action—and especially the Education Department’s botched rollout of the new FAFSA—all likely played a role.

Childcare and Apprenticeship
A new California apprenticeship grant offers childcare support to help mothers train as welders, electricians, and carpenters, Larry Gordon reports for EdSource. Recipients in pre-apprenticeship programs can get up to $5K a year for childcare, with full apprentices receiving up to $10K. The money is part of a new strategy after past efforts to diversify the trades showed little progress, with more funding going to recruiting campaigns.

Community College Partnerships
Federal investments can help create and deepen partnerships between workforce boards and two-year colleges, write New America’s Shalin Jyotishi and Morgan Polk, who point specifically to a former training fund that funneled $1.9B to 700 community colleges to equip unemployed and underemployed adults with industry-aligned credentials. Industrial policy investments could be another forcing mechanism for these partnerships.

Career Exploration
A third of K-12 students say they started thinking about careers in grades 6–8, with another 30% saying they started in grade 9–10, found a survey of 15K students by Pearson. Given deepening labor shortages across the U.S. economy, businesses should be helping K-12 students to consider their career opportunities and to better understand educational requirements and skill needs, writes Pearson’s Casey Welch.

Cloud Training
AWS announced it will give away up to $100M worth of its cloud services over the next five years to educational institutions that are working to prepare underserved students for in-demand jobs. Selected organizations will be able to use “cloud credits” from the company to offset the costs of creating AI assistants, mobile apps, coding curriculum, and the like.

Thanks for reading. Let me know what we missed? —PF