Healthcare & Social Assistance
22% of workforce
Service
Risk: Low
Physical delivery, human presence, and regulatory complexity insulate most roles; administrative roles are moderate exposure.
Where Americans Work
The Work Contract Is Broken
Work is no longer a single story. It's a map of industry concentration, collapsing contracts, geography, gender, and — increasingly — AI exposure that is reordering entire sectors at speed. The question isn't whether disruption is coming. It's whether anyone is building a floor beneath the people it displaces.
“For generations, work has been more than a paycheck — it has been a source of honor and dignity. When the modern workplace strips that away, people don't stop needing it. They find it elsewhere. The next wave of economic progress will be powered by those who rewrite the broken rules to bring back dignity.”
9.3M
Americans holding multiple jobs simultaneously
BLS, Nov 2025 — highest since tracking began in 1994
3.9 yrs
Median job tenure
Lowest since 2002
2.5 yrs
Half-life of a technical skill
Down from a decade
45%
Share of Gen Z holding a traditional full-time job
Down from a generational baseline near 70%
The Old Deal
What Work Used to Promise
For fifty years, an implicit contract organized American working life: show up, stay loyal, develop a skill, retire with dignity. The contract was never universal — it excluded most women, most people of color, most gig-adjacent workers — but it was real enough to structure a civilization's expectations.
The contract is gone. Median job tenure is 3.9 years. Only 45% of Gen Z holds a traditional full-time job. The skill they trained for expires in 30 months. The gig economy replaced the employer but not the power structure. And the AI wave is erasing the entry-level rungs that used to be how everyone got started.
The Old Deal
- —Lifetime employer, defined benefits
- —Skills with 10-year horizons
- —Median tenure: 15+ years (1983)
- —Entry-level → journeyman → expert ladder
- —Benefits: healthcare, retirement, paid leave
- —Work as identity anchor
The New Reality
- —3.9-year median tenure (2025 low)
- —Skill half-life: ~2.5 years
- —9.3M workers holding multiple jobs
- —Entry-level roles: first to be automated
- —70M gig workers, avg. $16.67/hour
- —57% of Gen Z has a side hustle (not by choice)
Employment Landscape
The Map of American Work
Every sector tells a different story about stability, compensation, AI exposure, and long-term viability. The largest employer in the U.S. is healthcare. The fastest-shrinking jobs in 2025 are in creative services, customer service, and entry-level tech.
Government & Public Administration
15%
Hybrid
Risk: Moderate
Frontline roles are insulated; administrative layers face meaningful AI substitution over 5–10 years.
Retail Trade
11%
Service
Risk: High
Checkout automation, AI inventory management, and e-commerce continue contracting employment; platform aggregators accelerating pressure.
Professional & Business Services
10%
Knowledge
Risk: High
Mid-level analysts, coordinators, and first-draft knowledge workers face the most acute near-term displacement.
Food Service & Accommodation
9%
Service
Risk: Moderate
Physical presence insulates delivery; ordering systems and back-of-house automation ongoing.
Manufacturing
8%
Trade & Craft
Risk: Moderate
Robotics automation is a 30-year trend; 7M jobs lost since 2000; remaining roles increasingly require technical skill.
Education Services
8%
Service
Risk: Moderate
Instructional roles more insulated; administrative and assessment-support roles have meaningful AI exposure.
Finance & Insurance
6%
Knowledge
Risk: High
Entry-level analysis, reporting, and advisory functions already contracting; Goldman, JPMorgan cutting junior analyst pipelines.
Construction
5%
Trade & Craft
Risk: Low
Physical craft, zero remote substitution; 349K new workers needed in 2026 alone, rising to 456K in 2027.
Transportation & Warehousing
5%
Hybrid
Risk: High
Waymo at 450K paid rides/week; autonomous trucking on a 5–8 year deployment horizon; UPS eliminated tens of thousands through automated hubs.
Technology & Information
3%
Knowledge
Risk: Critical
46% of code now AI-generated; entry-level developer employment (ages 22–25) fell nearly 20% from 2022 peak; advertising and PR shed 54,000 positions in 2024.
Skilled Trades (Electrical, HVAC, Plumbing)
4%
Trade & Craft
Risk: Low
Structurally irreplaceable; AI data center construction boom is ironically the single largest driver of trades demand; top-decile electricians earn $106K.
Interactive
AI Risk by Sector
Sort by workforce share or AI displacement risk. Filter to a single risk tier to see how the labor market clusters.
Sort by
Filter risk
Low risk
Moderate risk
High risk
Critical risk
Interactive
Workforce Composition by Work Type
Each bar shows how a category of work is internally split by AI displacement risk. Hover a risk tier in the legend to highlight that slice across all categories.
Trade & Craft
Physical craft and skilled trades — structurally hardest to automate.
17% of workforce
9%
8%
Knowledge
Analysis, code, and judgment work — AI's primary substitution zone.
19% of workforce
16%
3%
Service
Care, hospitality, and frontline service — physical presence insulates delivery.
50% of workforce
22%
17%
11%
Hybrid
Mixed physical/administrative roles — bifurcating exposure across the org chart.
20% of workforce
15%
5%
Workforce shares are approximate sector totals (BLS, 2025). Risk tiers reflect near-term AI displacement exposure across each sector's typical role mix.
Interactive
The Compressing Half-Life of a Technical Skill
In 1985 a technical skill held its value for roughly a decade. By 2025 that window has collapsed to about thirty months — and the curve is still bending. Hover any point for the context behind that year.
10 yrs
Skill half-life, 1985
2.5 yrs
Skill half-life, 2025
−75%
Durability lost in 40 years
~30 mo
Window before retraining
The Great Rebalancing
The Trades Are Back. The Question Is Whether Anyone Is Ready.
Knowledge & Service Work
~58% of workforce
- —Degree-heavy, urban-concentrated
- —AI's primary target zone for displacement
- —The “missing ladder rung” problem: AI now handles the entry-level tasks that used to train junior workers
- —Entry-level developer employment down 20% from 2022 peak
- —Advertising & PR: 54,000 jobs lost in 12 months
- —Customer service: 80% theoretical displacement risk; Salesforce cut 4,000 roles
- —The 56% wage premium for workers with AI skills is real — but only for those who make the transition
Trade, Craft & Physical Work
~42% of workforce
- —Historically undervalued, now structurally scarce
- —Median electrician: $62,350; plumbers: $62,970; HVAC: ~$60,000
- —Top-decile electricians: $106,000
- —Vocational enrollment up 16% YoY (2023) and 14% (2024) — consecutive double-digit growth
- —HVAC enrollment surged 41% from 2020 to 2024
- —55% of Gen Z say they’d consider a trades career — two-thirds credit social media
- —41% of construction workforce reaches retirement age by 2031 — the supply crisis is structural, not cyclical
- —The AI boom is the trades’ unexpected ally: data center construction requires electricians, HVAC specialists, and construction workers AI cannot replace
An HVAC technician finishing a 2-year program at 22 with zero debt is, on a net-worth-at-30 basis, frequently ahead of a 4-year graduate with $40,000 in debt — even if the nominal salary is higher.
Place as Destiny
Where You Live Determines What Work Is Available — and That Gap Is Widening
Urban / ~31% of workers
94% of U.S. job growth since 2000
Highest wages, densest professional networks, thickest labor markets. SF GDP per capita: $117,050 — nearly double the national figure. Top 10 metros produce 24.8% of national GDP. Remote work rate: ~32%. Dominant sectors: Tech, Finance, Media, Professional Services. AI exposure: highest of any geography.
Honest tensionUrban knowledge workers face the most acute AI displacement risk, even as urban labor markets remain the most dynamic overall.
Suburban / ~55% of workers
Primary beneficiary of remote work era
Access to urban labor markets at lower cost. 48% of millennials now live in suburbs. Remote work stabilized the suburban employment economy. Dominant sectors: Healthcare, Retail, Education, Construction. AI exposure: mixed — trades and care sectors insulated; administrative and mid-level professional roles exposed.
Honest tensionSuburban workers face the broadest range of AI exposure — from well-insulated trades to highly exposed administrative and service roles.
Rural / ~14% of workers
47% of rural counties have lost jobs since 2000
Rural median household income $66,600 vs. $80,600 urban — a 25% gap unchanged for two decades. Transfer payments now represent nearly $1 in $4 of rural personal income. Dominant sectors: Agriculture, Manufacturing, Government, Healthcare. AI exposure: lower on knowledge dimensions, but ongoing manufacturing automation continues eroding remaining industrial employment.
Honest tensionRural America has already absorbed a generation of deindustrialization. The AI wave's primary rural risk is in logistics and light manufacturing, not knowledge work.
Remote-work footnoteBefore COVID, roughly 5% of workdays were performed at home. That figure stabilized at 27% and shows no sign of reverting. 58% of those who relocated moved to suburbs of the same city; 29% moved to smaller cities; just 4% went rural. The remote work migration is not a rural renaissance. It is a mid-tier surge.
Gender and the Labor Map
Gender Predicts Which Sector You're In — and Which AI Risks You Face
Gender remains one of the strongest predictors of which sector a worker occupies — and therefore which disruption risks they face. The patterns are structural, not accidental: they reflect decades of credentialing, caregiving norms, and occupational sorting that AI is now intersecting with in distinctly gendered ways.
Women — Sector concentration
- Healthcare services~77% female
- Education~73% female
- Social services~68% female
- Administrative / office support~72% female
- Customer service~65% female
AI implication: Customer service (80% theoretical displacement risk) and administrative work are where measurable job loss has already begun at scale. Salesforce cut 4,000 roles. Women carry disproportionate exposure here. Healthcare and education remain more insulated — physical delivery and human judgment anchor those roles.
Structural note: Women earned approximately 82.7 cents per dollar men earned in full-time employment (U.S. Census, 2023) — a gap that widened for the first time in 20 years. When AI-driven displacement hits the sectors women are concentrated in, the wage gap compounds into a wealth and stability gap.
Men — Sector concentration
- Construction~88% male
- Transportation~78% male
- Technology / Software~74% male
- Manufacturing~71% male
- Finance / Insurance~55% male
AI implication: Men disproportionately occupy both the trades (low AI risk, high demand) and senior tech roles (AI-augmented). But entry-level tech — where young men concentrate — faces the sharpest near-term displacement. Developer employment ages 22–25 fell nearly 20% from its 2022 peak. Goldman Sachs found U.S. tech workers in their twenties in AI-exposed occupations saw unemployment rise 3 percentage points in the first half of 2025 alone.
The male formation collapse: 1 in 5 men aged 25–54 is neither working nor looking for work. Manufacturing has shed 7M jobs since 2000. The three historical formation structures for working-class men — military service, trade apprenticeship, stable industrial employment — have all declined simultaneously. Work wasn't just income. It was identity, community, purpose.
The Disruption Horizon
Not Every Job Is Equally at Risk — and the Gap Is Widening Fast
AI is not eliminating jobs uniformly. It is hollowing the middle — automating the journeyman layer, eliminating entry-level training grounds, and concentrating remaining value at the extremes of skill and physicality. The result is a labor-market barbell: high-value expert work on one end, irreplaceable physical/craft work on the other, and a shrinking, precarious middle.
Left end
Trades · Craft · Physical
Low AI risk · rising demand · shortage economy
Shrinking middle
Mid-Level Knowledge
Highest displacement density
Right end
Senior Expert · Judgment · Relationship
AI-augmented, not replaced — requires irreplaceable context
52%
of AI workplace use is augmentation; 45% is automation — trending toward more automation (Anthropic Economic Index, 2026)
46%
of all code now written by AI tools (GitHub Copilot)
80%
theoretical displacement risk for customer service roles
56%
wage premium for workers with demonstrated AI skills (PwC, 1B+ job postings)
92M / 170M
WEF projection by 2030: jobs displaced / created — but the displaced and the created will not be the same people
$8.9T
estimated annual global cost of disengagement in lost productivity (Gallup, 2025)
The invisible displacement
Yale Budget Lab's most important finding: AI appears to be suppressing hiring more than destroying existing jobs. A generation may fail to build career capital not because they lost jobs, but because the jobs that would have existed simply never appeared. This is the “missing ladder rung.”
The deskilling paradox
Research shows Anthropic's Claude tends to handle the higher-education components of jobs, producing a net deskilling effect. The tasks AI absorbs are often the most skilled parts of the role. Short-term productivity gains may come at the cost of long-term workforce capability — a paradox with compounding consequences for whoever is starting their career right now.
What This Means
Work Was Never Just a Paycheck
Work, for most of human history, was not a source of identity. It was a condition of survival. The 20th-century innovation was to transform it into something more — a source of meaning, community, upward mobility, even self-actualization. That transformation was never universal. It excluded most women, most people of color, most of the Global South. But it was real enough to structure an entire civilization's expectations.
Gen Z inherits the expectations without the infrastructure. They've been told that work should be meaningful, but the meaningful jobs don't pay. They've been told to build skills, but the skills expire before the student loans do. They've been told to be entrepreneurial, but the platforms take 20–42% of what they earn. 69% live paycheck to paycheck.
The trades worker who builds the data center that runs the AI that displaces the analyst one floor up is not a footnote. He is the economy. The nurse's aide who cannot be automated — whose work is irreducibly, stubbornly human — is not a consolation prize. She is what the future of work actually looks like, if we're honest about it.
Dignity is not a reward for productivity. It is the precondition for it. The next wave of economic progress will be powered by those who rewrite the rules to restore it.
“For generations, work has been more than a paycheck — it has been a source of honor and dignity. When the modern workplace strips that away, people don't stop needing it. They find it elsewhere. The next wave of economic progress will be powered by those who rewrite the broken rules to bring back dignity.”
Related Research