Some investors are betting AI will slash working hours and supercharge leisure. The evidence—and the caveats—suggest a slower, uneven transition.

LONDON/NEW YORK

— A decade ago, the idea that most people might work just three days a week belonged to utopian Twitter threads and techno-optimist conference talks. Today it is part of real investment theses. A small but vocal set of financiers and operators argue that artificial intelligence will compress routine work, push productivity sharply higher and, in the process, expand the market for leisure. Their bets are no longer theoretical: live-events roll‑ups, gaming deals and new entertainment platforms are being financed on the assumption that more free time—and disposable income—are coming.

The most explicit wager comes from Hollywood power broker Ari Emanuel. In early October he unveiled MARI, a nearly $3 billion vehicle that has already snapped up high‑profile assets from tennis tournaments to art fairs. The logic, laid out to prospective investors, is plain: if AI helps many white‑collar jobs consolidate into fewer days, demand for premium, in‑person experiences will surge. Others share different versions of the same thesis—private‑equity shops amassing sports and media assets, and strategics doubling down on games—not because of streaming wars fatigue, but because they expect a structural reallocation of hours from desks to downtime.

It is a seductive vision that recalls the long arc of the 20th century, when technological progress slowly shortened the workweek. But the leap from today’s frenetic knowledge economy to a three‑day norm rests on three questions: Will AI deliver a step‑change in productivity at the macro level? Will the gains be widely shared? And even if they are, will people and employers actually choose fewer hours over higher pay and output?

A productivity promise, still uneven in the data

Corporate memos and strategy decks increasingly treat AI as a cost‑saver and throughput multiplier. Goldman Sachs this year reiterated estimates that widespread adoption of generative systems could lift productivity levels in advanced economies by roughly 15 percent once fully embedded—an effect large enough to alter trend growth and, during the transition, briefly raise unemployment as tasks are reallocated. That promise is starting to reshape staffing plans: on October 14, Goldman executives told employees the bank would slow hiring and consolidate roles as it leans into AI tooling across client onboarding, reporting and operations—while still expecting headcount to rise by year‑end as resources shift to higher‑value work.

Outside finance, software and services firms report mixed but improving gains. Some teams have shaved minutes off routine tickets, drafts or analyses; others complain about a flood of mediocre output that needs human clean‑up. The difference often comes down to workflow design and data quality. In aggregate statistics, the big step‑change hasn’t arrived yet. But the direction of travel—in particular in coding, customer support, and content localization—has been enough for investors to assign real weight to the leisure thesis.

The four‑day frontier, from pilots to policy

If a three‑day week is the eye‑catching headline, the nearer‑term beachhead is four. A rolling series of coordinated pilots across the UK, Portugal, Germany and beyond have found consistent improvements in worker well‑being alongside stable—or sometimes higher—output. In the UK’s large‑scale trial, most participating companies kept the policy after six months; self‑reported burnout fell sharply, attrition slowed and revenues nudged higher. A meta‑pattern is emerging: compressing hours forces clearer prioritization and better meeting discipline, and the time dividend shows up as lower sick days and steadier teams.

Still, translating pilots into a nationwide norm is a different proposition. Sectors with continuous operations—healthcare, logistics, hospitality—can’t simply remove a day without backfilling shifts or automating aggressively. Even in desk jobs, a shorter week depends on target‑setting and customer expectations: when everyone’s off on Friday, response‑time anxiety eases; when schedules diverge, it doesn’t. That is partly why governments, not just employers, loom large over the pace of change. Without nudges—tax treatment, overtime rules, civil‑service leadership—adoption will be patchy.

Winners, losers and the leisure supply chain

If hours do fall, money doesn’t necessarily follow immediately. A typical working household may swap some streaming time for lower‑cost outdoor activities rather than premium sports tickets. Leisure’s winners would likely cluster where AI cannot easily duplicate the experience: live performance, travel, community sport, boutique fitness, maker spaces and hands‑on learning. Conversely, parts of the attention economy could face diminishing returns as AI saturates supply; lowering the cost of content creation risks further commoditization unless paired with scarce live or social components.

For capital allocators, the playbook being assembled is surprisingly old‑fashioned: acquire durable IP and scarce venues; own the dates on the calendar that people plan around; build pricing power via membership and bundles; and spread the fixed costs of production over global franchises. What’s new is the timing argument—front‑running a work‑time shift that would expand the total addressable market over the next five to ten years.

Distribution matters more than invention

Even if AI lifts the productivity pie, who gets the extra slice will decide whether the average person works less. In the post‑war decades, unions and policy helped convert efficiency gains into shorter hours and rising wages. In the 2000s and 2010s, the benefits of digitization accrued disproportionately to firms and skilled workers, while overall hours in the U.S. barely budged. Today’s early evidence points both ways: some employers are reinvesting AI savings into customer service or growth, others are trimming roles and freezing backfills. Without broader bargaining power or policy intervention, a three‑day week risks remaining a perk for elite teams rather than a baseline for the median worker.

Signals to watch

Three indicators will show whether the age of leisure is truly arriving or just an investable narrative:

1) Macro productivity and hours worked

Quarterly productivity growth in services, coupled with average weekly hours in advanced economies, will reveal whether AI is bending the curve. Sustained gains above pre‑pandemic trends—without a surge in overemployment—would support the case for shorter schedules.

2) Policy diffusion

Look for city and national governments trialing shorter weeks in public services, and for legal frameworks that make four‑day schedules the default for certain roles. Public‑sector adoption often catalyzes private‑sector imitation.

3) Enterprise architecture

The extent to which companies rebuild processes around automation—structured data pipelines, human‑in‑the‑loop quality gates, retrieval‑augmented knowledge bases—will determine whether AI removes work or just moves it. Firms that redesign end‑to‑end workflows, not just bolt tools onto old habits, capture the hours back.

Bubbles, brakes and the likely landing zone

Experienced investors say narratives tend to overshoot before reality catches up. There is a non‑trivial risk that some leisure‑driven deals are pulled forward on optimistic assumptions about time use, only to meet a consumer who is richer in tools but not in hours. On the other hand, many white‑collar teams already operate with implicit quiet Fridays or asynchronous Mondays—proto four‑day weeks by another name. The path most consistent with current evidence is evolutionary: more four‑day arrangements, concentrated in professional services and tech‑adjacent fields; expanding live‑experience markets in cities with high adoption; and continued debate over how to equitably share the gains.

The three‑day week is not imminent, and may never be universal. But AI is finally strong enough to force a renegotiation of work’s boundaries. Whether that renegotiation produces an age of leisure—or simply another age of hustle—will depend less on what models can do than on what employers, policymakers and workers decide to do with what models make possible.

Sources

• Ari Emanuel’s MARI leisure thesis and funding details reported in the Financial Times, October 2025; and related coverage. • Goldman Sachs research (Aug. 2025) estimating ~15% productivity level lift from generative AI when fully adopted; and an Oct. 14, 2025 internal memo describing AI‑driven operating changes. • Results from multi‑country four‑day week trials via Autonomy/4 Day Week Global and recent summaries in Scientific American and APA reporting.