AI and the Future of Work Roundtable

As part of NGA's Working Group on AI and the Future of Work, Governors' AI and workforce advisors met last week to share best practices and hear from national experts, including NVIDIA co-founder Chris Malachowsky.

The session started with State and Territory leaders discussing the efforts that they are leading on behalf of their Governors to prepare for the significant impact that advancements in AI will have on the labor market. Common strategies include real-time dashboards tracking occupational exposure, rapid upskill scholarships, business-led strategies that gather intelligence directly from employers, and public-private partnerships that co-design curriculum with industry. Several states are working to expand AI literacy beyond government agencies to local officials, small businesses, and the general public, treating accessibility as a prerequisite for any broader strategy to work. Others described a more fundamental challenge: local businesses and communities don't just lack AI capacity, they often can't articulate what they need.

The session surfaced as many questions as answers: How do you explain the difference between AI and automation to legislators proposing guardrails? Should skill-matching infrastructure be built nationally rather than state by state? Are unemployment trust funds the right buffer for displaced workers? And with AI evolving faster than most task forces can keep up, how do you keep your strategies from going stale before the ink is dry? It is clear that the Working Group has an important task in front of them.

McKinsey & Company then took attendees through new research on the impact of AI on the labor force at the national and state/territory level. According to their research more than half of all U.S. work hours are theoretically automatable using technologies available today. AI will not eliminate jobs en masse, it will permeate virtually every occupation, shifting which tasks humans perform rather than removing the human entirely.

State workforce systems face two simultaneous demands: equipping workers with the AI tools they need to succeed tin today's economy, and enabling transitions across occupational categories as some shrink and others grow. Demographic decline sharpens the stakes: AI productivity gains are among the few forces large enough to offset the GDP drag of an aging population, but only if human capacity is redeployed rather than simply reduced.

The ensuing discussion identified several priorities: tighter coordination across training providers and employers around a shared North Star; governance frameworks defining each actor's role in AI adoption; a shift from credential-counting to productivity-based performance measures; and new apprenticeship models that pair younger workers' AI fluency with the judgment of experienced professionals. State-level occupational data should anchor all of it as the automation footprint looks different in every jurisdiction.

Chris Malachowsky, co-founder of NVIDIA, closed the roundtable saying that the window for cautious observation has closed. "It's coming so fast," he said. "The innovations, the risks, and the rewards are going to smack you in the head." His central argument was a reframe he urged every advisor to carry back to their capitals: AI is not an app or a trend. It is infrastructure, as essential as water and electricity. States and territories should not treat it as a novel technology and start investing in it the way they fund protected public systems that don't require short-term payback. That means the full stack: models, data centers, and the energy to power them.

His closing message was simple: don't wait for a perfect plan. States with rural communities and workforce gaps stand to gain the most, and the fundamentals: critical thinking, good judgment, asking smart questions, are not new, and can start in kindergarten. "Come up with a strategy. Get going on it yesterday."