Meta Chief Data Officer: AI May Offset Global Population Collapse

Meta chief data officer Alex Schultz warned in remarks reported by the Financial Times on November 5 that the world faces a potential "population collapse," and he argued that artificial intelligence could supply the productivity gains economies will need when the pool of working-age adults starts shrinking. His argument inverts the most common critique of AI in the workplace: instead of destroying jobs people need, the technology may become essential to filling jobs that would otherwise go undone.

Schultz, who serves as chief data officer at Meta Platforms (NASDAQ:META), pointed to projections suggesting the global population could peak as early as 2064. His concern is not simply that there will be fewer people. It is that there may be too few working-age adults to support growing numbers of retirees as birth rates fall and people live longer.

What do the demographic numbers show?

The Congressional Budget Office's January 2026 demographic outlook quantifies the problem for the United States. It projects that the number of working-age adults between 25 and 64 for every person aged 65 or older will decline from 2.7 today to just 2.2 by 2056.

That shift means a shrinking pool of workers supporting an expanding retirement population. Schultz believes AI productivity gains could bridge the gap. Businesses might produce more goods and services with fewer employees, allowing economies to maintain living standards despite slowing population growth.

The CBO also projects that America's population will stop growing around 2056 and begin declining afterward. An aging society will need to accomplish more with a smaller workforce, and that structural fact will not disappear when today's AI investment boom eventually cools.

The context matters for the AI debate. Companies are investing hundreds of billions of dollars in technology that promises to accomplish more with fewer employees, and workers worry whether their jobs will survive. Schultz's framing, coming from a top executive at one of the world's biggest AI investors, casts the same technology as a response to a separate economic threat: falling birth rates and aging populations that could leave many countries struggling to find enough workers to support their economies.

Is AI already changing the labor equation?

Schultz is not speaking entirely in hypotheticals. According to the Financial Times, Meta already uses AI agents to resolve more than 10% of certain data-warehouse monitoring tasks without human intervention. He also cited improvements in software development and legal work.

Broader economic figures offer intriguing, though not conclusive, support. Recent Bureau of Labor Statistics productivity data showed American businesses producing more output without a corresponding increase in working hours. Productivity growth is beginning to resemble the economic transformation AI companies have promised, although the government figures do not establish how much of the improvement comes from AI.

The implications extend beyond technology companies:

• Hospitals facing nursing shortages could use AI to reduce paperwork.
• Manufacturers could automate routine inspections.
• Financial institutions could handle more transactions without expanding administrative staff.

None of those applications replaces the need for nurses, factory technicians, or financial professionals. They potentially allow existing employees to accomplish more.

Yet there is a contradiction investors should not overlook. The Federal Reserve's Beige Book has already documented employers using AI to limit hiring. Productivity gains may help address tomorrow's labor shortages while creating difficult employment transitions today. The technology many fear will destroy employment could eventually help compensate for a shortage of workers — but the two effects will operate simultaneously, not sequentially.

What problem can AI not solve alone?

Greater productivity does not automatically translate into greater prosperity for everyone. If AI allows businesses to produce more with fewer employees, shareholders could benefit from lower costs and higher profits. Displaced workers may struggle to find comparable employment, particularly if the gains concentrate among a handful of technology companies.

The difference becomes critical when measured against Schultz's demographic warning. A country with fewer workers needs greater output per worker. It also needs households with enough income to purchase what businesses produce. AI cannot independently solve that distribution problem.

Nor does every industry benefit equally. Software can accelerate accounting and administrative work, but caring for an aging population still requires nurses, caregivers, facilities, and physical infrastructure. The sectors where the demographic squeeze will bite hardest are, in significant part, the sectors that automation addresses least directly.

There are encouraging signs on the other side of the ledger. AI adoption can accompany business formation and new employment opportunities. Whether those opportunities offset displaced jobs remains uncertain.

Why does the argument matter for investors?

For investors, the longer-term opportunity may extend well beyond the companies developing AI models. Businesses that successfully apply automation to genuine labor shortages could enjoy productivity advantages for decades.

Schultz's warning highlights a demographic problem that will outlast the current AI investment cycle. The CBO's 2056 inflection point gives the issue a fixed horizon. The projection that the working-age support ratio falls from 2.7 to 2.2 over three decades defines the scale of the adjustment the U.S. economy must make.

Investors should look beyond promises of replacing employees and focus on companies using AI to expand output, improve services, and solve measurable labor shortages. Those benefits could prove more durable than the latest AI spending forecasts, and they rest on demographic projections rather than model releases. Higher productivity alone, however, will not guarantee broad economic prosperity — a constraint Schultz's argument acknowledges only implicitly, and one that policy rather than software will have to address.