BusinessMarkets, money & strategy
Meta data chief Alex Schultz says falling birth rates make AI productivity gains vital
Meta chief data officer Alex Schultz told the FT's Future of AI summit that falling birth rates make AI-driven productivity gains essential. The demographic trend is well documented, but evidence that AI delivers the needed productivity remains mixed, including at Meta.
From the Financial Times · AI article of November 4, 2026 and cited sources: Schultz argued that AI productivity gains will be needed because there will be fewer people. [1] [3] [4]
SOURCES[1] ‘Population collapse’ makes AI productivity gains vital… ft.com[3] UN News: World Population Prospects 2024 news.un.org[4] Fortune: McKinsey on population decline fortune.com
AT Pulse research deskPublished · Updated · 4 min read · AI-assisted analysisAI-assisted, not human-edited. Methodology Corrections Who owns this site
- What happened
- Schultz argued that AI productivity gains will be needed because there will be fewer people. [1] [3] [4]
- Original source
- Financial Times · AI article, independent perspective
- Brief by
- AT Pulse research desk AI-assisted; publisher and contact on the desk page
- When
- November 4, 2026 article published
- Covered
- Nov 4 edition
- Evidence
- 9 sources across 9 sites
- 1‘Population collapse’ makes AI productivity gains vital…ft.com
- 2The Next Web: Meta planned to cut some teams by 60%thenextweb.com
- 3UN News: World Population Prospects 2024news.un.org
- 4Fortune: McKinsey on population declinefortune.com
- 5Mediaweek: Meta names Alex Schultz Chief Data Officermediaweek.com.au
- 6Pragmatic Engineer: Meta wanted to reduce teams by 60%blog.pragmaticengineer.com
Reach: not measured yet
Schultz argued that AI productivity gains will be needed because there will be fewer people. The demographic premise has strong support. UN projections show more than half of countries below replacement fertility and 63 already past peak population. McKinsey estimates that ageing economies need productivity growth two to four times current rates, or longer working lives, to keep living standards steady. But global population is still projected to grow until the mid-2080s, so the decline is regional rather than worldwide. [1] [3] [4]
About ten weeks before the speech, Reuters reported that Meta had shelved part of a plan to shrink some teams by up to 60% using AI. Schultz was among the executives who studied AI-native startups for the plan. According to leaked internal data, code output rose 220% while user-facing features rose only 36%, and major incidents rose 40%. Meta cut about 8,000 jobs in May but cancelled a second wave. The company says it never intended to cut 60% of the whole company. [2] [6]
Output volume is a poor proxy for productivity. In METR's randomized trial, experienced open-source developers using early-2025 AI tools took 19% longer on tasks while believing they had been about 20% faster. Teams should track delivered outcomes and incident rates before resizing around AI. [8] [2]
Ageing populations create real productivity pressure. Describing AI as filling labour gaps rather than replacing workers may shape how firms justify job cuts.
Read the full assessment
The economy-wide gains required have not been shown, and coding speedups alone would not reach them.
Sources are in hand. Here is the dossier.
Executive brief
Alex Schultz, Meta's chief data officer, told the FT's Future of AI summit that falling birth rates mean AI will be needed to support a shrinking population (FT). The surprising part is the timing. Just ten weeks earlier, Reuters reported that Meta had dropped an internal plan to shrink some teams by up to 60% using AI. Internal data showed AI agents producing far more code without a matching rise in shipped features (The Next Web). The demographic facts are solid. Whether AI can deliver the productivity they demand has not been shown.
What changed and event timeline
UN confirms the demographic turn
World Population Prospects 2024 found more than half of countries below replacement fertility, and 63 countries, including China, Germany and Japan, already past peak population ().
McKinsey sets the productivity bar
McKinsey Global Institute said countries with ageing populations need productivity growth two to four times current rates, or longer working lives, to hold living standards ().
Meta's "Project OT."
The plan started in January. In May, Meta cut about 8,000 jobs (10% of staff) and moved 7,000 people into AI roles. It cancelled a second wave planned for November ().
Schultz becomes chief data officer
He moved from CMO to a newly created post. He said AI creates "effectively unlimited demand for insights," and Denise Moreno became CMO ().
Reuters exposes Project OT
The report says Schultz and product head Naomi Gleit studied AI-native startups in Asia and piloted similar structures inside Meta ().
"Population collapse" remarks
On the summit's opening day in London, Schultz argued AI productivity is vital because there will be fewer people (;).
Capabilities and access
This story is a policy and strategy argument. No model, product or version was announced. Only the FT excerpt's one-line summary of the remarks is available. No further independent coverage of the speech turned up, so its detailed claims cannot be checked.
Technical analysis for researchers and developers
- What Meta measured internally. Counting lines of code overstated the gains.
- In METR's randomized trial, 16 experienced open-source developers using early-2025 AI tools took 19% longer on tasks, yet believed they had been about 20% faster (METR).
- What this means for practice. Measure delivered outcomes and incident rates, not output volume.
Read the full section
- What Meta measured internally. According to Reuters' account of internal data, code output rose 220% but user-facing features rose only 36%. Major technical and security incidents rose 40%, and staff time spent firefighting them rose 70% (The Next Web). Counting lines of code overstated the gains.
- Independent evidence. In METR's randomized trial, 16 experienced open-source developers using early-2025 AI tools took 19% longer on tasks, yet believed they had been about 20% faster (METR).
- What this means for practice. Measure delivered outcomes and incident rates, not output volume.
Claims and evidence
- Falling birth rates make AI vital: This is the speaker's argument (FT). The demographic premise is backed by the UN and McKinsey.
- Workers per retiree fall from 6.5 in 2025 to a projected 3.9 by 2050 (McKinsey, via Fortune).
- Project OT figures come from leaked internal data. Meta says it "never intended to cut 60% of the whole company" (The Next Web).
Read the full section
- Falling birth rates make AI vital: This is the speaker's argument (FT). The demographic premise is backed by the UN and McKinsey. The claim that AI is the answer is not.
- Workers per retiree fall from 6.5 in 2025 to a projected 3.9 by 2050 (McKinsey, via Fortune).
- Project OT figures come from leaked internal data. Meta says it "never intended to cut 60% of the whole company" (The Next Web).
Context and prior work
- In 2025, Schultz described three ways AI affects headcount: doing existing work faster, making new kinds of work possible, and making once-too-costly work, such as support bots, affordable (AOL/Business Insider).
- The UN projects that people aged 65 and over will outnumber those under 18 by the late 2070s.
Read the full section
- In 2025, Schultz described three ways AI affects headcount: doing existing work faster, making new kinds of work possible, and making once-too-costly work, such as support bots, affordable (AOL/Business Insider).
- The UN projects that people aged 65 and over will outnumber those under 18 by the late 2070s. It also projects that 63 countries already past their peak will lose 14% of their population over 30 years (UN News).
Limitations, safety and contested findings
Read the full section
- "Population collapse" overstates the global picture. The UN projects world population still rising to about 10.3 billion in the mid-2080s, with fast growth in places like Niger and Somalia (UN News). Decline is regional.
- The workforce effects are disputed. Project OT led to an employee backlash: Meta's favorability score in its staff survey fell from 74% to 55%, and 26 employees sued (The Next Web).
- The productivity evidence points both ways. METR cautions that its results are a snapshot of early-2025 tools and may not generalise (METR).
Business and practitioner implications
- Describing AI as filling a demographic gap rather than replacing workers is likely to shape how big platforms talk about cuts.
- McKinsey's two- to fourfold productivity bar is economy-wide. Coding speedups alone do not reach it.
- Before resizing teams, pilot with outcome metrics, incident tracking and supervision of AI agents. Meta's experience shows what happens without them.
Read the full section
- Reframing. Describing AI as filling a demographic gap rather than replacing workers is likely to shape how big platforms talk about cuts. Read it alongside Meta's actual staffing decisions (Pragmatic Engineer).
- Scale of the target. McKinsey's two- to fourfold productivity bar is economy-wide. Coding speedups alone do not reach it.
- For managers. Before resizing teams, pilot with outcome metrics, incident tracking and supervision of AI agents. Meta's experience shows what happens without them.
The source trail.
Sources (9)
‘Population collapse’ makes AI productivity gains vital, says top Meta executive
Feed excerpt only.
ft.com