Note: This is a research note supplementing the book Unscarcity, now available for purchase. These notes expand on concepts from the main text. Start here or get the book.
The 2025-2030 Labor Cliff: The Great Unbundling of Human Work
Try explaining to a Medieval peasant that one day, machines would do all the farming, and 97% of people would have to find something else to do. They’d probably ask, “Then who eats?”
We’re that peasant now. Except instead of tractors replacing farm hands over 200 years, it’s AI replacing knowledge workers over ten. And unlike the peasant, we have spreadsheets, so we can watch it happen in real time.
The numbers arrived in 2025, and they were bad. The 2026 numbers are worse.
The Crime Scene: 1.2 Million in 2025, and AI Is Now the #1 Reason
In 2025, U.S. employers announced 1.2 million layoffs, an increase of 58% over 2024 and the highest annual total outside of COVID since the Great Recession. Q4 2025 saw the highest quarterly layoffs since 2008. Year-to-date hiring fell to its lowest since 2010.
We weren’t in a recession. GDP was positive, inflation was cooling, and corporate profits sat at record highs. Yet companies were firing workers at financial-crisis rates.
The reason is that they’re substituting labor, not just cutting costs. Every company announcing layoffs is simultaneously announcing billions in AI investment. Microsoft cut 15,000 workers and invested $80 billion in AI infrastructure. Meta fired 4,200 people and redirected the savings to training Llama. Amazon eliminated 14,000 corporate jobs in 2025, then announced another 16,000 in early 2026.
The companies call it “restructuring.” What it amounts to is replacement.
Then March 2026 broke the dam. For the first time in Challenger, Gray & Christmas’s history of tracking U.S. job cuts, artificial intelligence was the #1 cited reason for layoffs, with 15,341 March cuts directly attributed to AI, 25% of the monthly total. February had only attributed 4,680 cuts to AI (~10% of total). The AI share tripled in a single month.
April 2026 confirmed the pattern. The April Challenger report logged 83,387 total cuts, with AI cited in 21,490 of them (26%), the second straight month AI was the top stated reason for layoffs. Two consecutive months above a quarter of all cuts is how a spike becomes a structure; AI has now been blamed for 49,135 cuts in 2026 alone. The names told the story. Meta announced 8,000 layoffs (10% of its workforce), raising 2026 capex guidance to ~$135 billion (an 87% YoY jump, mostly AI infrastructure). Microsoft offered buyouts to ~7% of U.S. staff (up to 8,750 people). Snap cited “rapid AI advancements” by name. And it had spread past the AI labs: PayPal and Ticketmaster both restructured around AI, dragging payments and ticketing into the cut.
May 2026 turned the pattern into a record. The May Challenger report logged 97,006 cuts - up 16% from April, the highest May since 2020, and a third straight monthly rise. AI was named in 38,579 of them, fully 40% of the month’s total: the biggest monthly AI figure Challenger has ever recorded and the third consecutive month AI topped every stated reason. Through May, AI has been blamed for 87,714 cuts this year (22% of the total), already surpassing all of 2025. The tech sector alone shed 38,242 jobs in May, its worst month since August 2024.
Then Meta showed what “substitution” looks like up close. When the cuts executed on May 20, Meta fired 8,000 people and reassigned another 7,000 into four new AI-native units, retitling staff “AI builder” and reorganizing around “superintelligence” pods. Same company, same week: thousands out the door, thousands redrawn around the machine. This is not cost-cutting. It is a workforce being rebuilt to point at AI instead of away from it.
June 2026 is the month that proved it’s structural. Total announced cuts collapsed to 45,849 — down 53% from May, the lowest month since December 2025, and 4% below June 2025. The wave receded. And AI’s share went up: 14,029 cuts, 31% of the total, leading every stated reason for a fourth consecutive month. Challenger has tracked AI as a distinct reason since 2023 and has never recorded a streak like it.
That divergence is the whole argument in one data point. Cyclical layoffs breathe with the economy; substitution layoffs don’t. When the volume halved and the AI share climbed nine points, what was left standing was the part that isn’t going to reverse when conditions improve.
Through six months: 101,743 cuts attributed to AI, ~23% of everything Challenger tracked — nearly double the 54,694 blamed on AI in all of 2025, in half the time. The technology sector shed 139,156 jobs in H1, up 83% year over year, and now accounts for about a third of all US job cuts.
Tom’s Hardware reported that 47.9% of Q1 2026 tech-sector layoffs were AI-attributed, roughly half the industry that built AI being cut by AI.
The Challenger Report: Reading Between the Lines
Challenger, Gray & Christmas tracks layoff announcements like a coroner tracks causes of death. Here’s the full-year 2025 picture:
| Metric | Figure | Translation |
|---|---|---|
| Total 2025 layoffs | 1.2 million | 58% higher than 2024 |
| Q4 2025 layoffs | Highest since 2008 | Financial crisis territory |
| YTD hiring (2025) | Lowest since 2010 | Not replacing workers |
| October 2025 alone | 153,074 | Highest October in 22 years |
| #1 reason cited | DOGE impact | 293,753 jobs (+20,976 downstream) |
| Government sector | 308,167 cuts | +703% vs. 2024 |
| #6 reason cited | AI | 54,694 jobs |
Notice that AI ranks sixth as a stated reason. But look at what ranks above it: “restructuring,” “economic conditions,” “closings.” What do you think companies are restructuring toward? They’re not rebuilding their org charts to hire more humans.
The stated reason and the real reason are doing a complicated dance. Companies say “efficiency”; they mean “algorithms.”
Source: Challenger Year-End 2025, CNBC, CNBC: January 2026
Tech Sector: Ground Zero
If you want to see the future of all white-collar work, watch what’s happening in tech. The sector that was supposed to be immune to automation is getting automated first.
The Hall of Layoffs (2025-2026)
| Company | Layoffs | What They’re Building Instead |
|---|---|---|
| Meta | 4,200 (2025) + 8,000 (Apr 2026) | $135B 2026 capex (87% YoY jump), AI infrastructure |
| Microsoft | ~15,000 (2025) + ~8,750 buyout offers (Apr 2026) | Copilot, OpenAI, AI data centers |
| Amazon | 14,000 (2025) + 16,000 (2026) | AWS AI, robotics |
| Intel | 21,000+ (~20% of workforce) | AI chips |
| Verizon | 13,000+ | Network automation |
| IBM | 8,000-9,000 | AI replacing HR and admin |
| Snap | 4,000+ (2026) | Cited “rapid AI advancements” by name |
In 2025, global tech layoffs reached nearly 245,000 workers, with ~70% from US-headquartered companies. In 2026, the pace has accelerated. By late April, over 92,000 tech workers had been laid off according to Layoffs.fyi, bringing the cumulative total since 2020 to nearly 900,000. The Challenger tech-sector total for Q1 2026 alone hit 52,050 cuts, the highest first-quarter figure since 2023. April 2026 added another ~40,000.
The irony is thick enough to cut: the people who built the machines are being replaced by the machines they built. By early 2026, 51% of all code committed to GitHub was either AI-generated or substantially AI-assisted, up from 46% just months earlier. For Java developers in Copilot-enabled environments, that figure hits 61%. The humans aren’t collaborating with AI; they’re being outpaced by it.
Source: Tom’s Hardware Q1 2026 Tech Layoffs, CNBC: Meta + Microsoft April 2026, BusinessToday: April 2026 Tech Layoffs
The BLS Numbers: 4.2% Unemployment, for the Wrong Reason
The Bureau of Labor Statistics’ March 2026 report showed 4.3% unemployment, nominally an improvement from February’s 4.4%. But the underlying movement is volatile, not healthy.
February 2026 shed 133,000 jobs (revised down from initial -92,000 estimate). March 2026 rebounded by +178,000, mostly healthcare resuming after February’s strike-driven dip. The unemployment-rate decline came largely from people leaving the labor force, not finding work.
April 2026 then held the rate at 4.3% for a third straight month, with payrolls up a thin +115,000 and labor-force participation sliding to 61.8%, the lowest since October 2021, as another 226,000 people stopped looking. Hiring in 2026 is averaging roughly 76,000 jobs a month against the ~150,000 needed just to keep up with population growth. The rate holds flat because the denominator is shrinking, not because the market is stable.
May 2026 (released June 5) briefly complicated the tidy collapse narrative: payrolls beat hard at +172,000 against an 80,000 forecast, unemployment held at 4.3% for a fourth straight month, and the BLS revised the prior two months up by 93,000.
That beat did not survive contact with the next report. June’s release (July 2) revised May back down to +129,000 and April down to +148,000, then printed +57,000 for June against a 115,000 consensus. Unemployment fell to 4.2% — the first move off 4.3% in five months — because 720,000 people left the labor force, dragging participation to 61.5% and the employment-population ratio to 59.0%. Leisure and hospitality shed 61,000 on weak seasonal hiring, erasing gains in professional services (+36,000), social assistance (+25,000), and health care (+22,000).
Two things are worth taking from that. First, the summer-job and seasonal-service channel — the escape hatch that has absorbed displaced workers for sixty years — did not open this year. Second, and more uncomfortable for anyone quoting these numbers: 2026’s initial prints have been revised down repeatedly. The optimistic reading is always available on release day and keeps getting withdrawn a month later.
What the Numbers Actually Show
| Metric | Rate | What It Means |
|---|---|---|
| Overall unemployment (Jun 2026) | 4.2% | Fell only because 720,000 people quit looking |
| Nonfarm payrolls (Jun 2026) | +57,000 | Half the consensus, a third of replacement pace |
| Labor force participation (Jun) | 61.5% | Down 0.3pp in one month; post-2021 low |
| Employment-population ratio (Jun) | 59.0% | The measure you can’t game, and it fell |
| Leisure and hospitality (Jun) | -61,000 | The seasonal entry-level channel didn’t open |
| Professional/business services (Jun) | +36,000 | Month’s largest gain |
| Health care + social assistance (Jun) | +47,000 | The economy’s last reliable engine |
| April & May revisions | -31,000 / -43,000 | Both months revised down after the fact |
| Real hourly earnings (YoY) | +3.4-3.5% | Near the lowest annual wage growth since May 2021 |
The trajectory looks like a heart-attack chart: October 2025 -105,000, November +64,000, February 2026 -133,000, March +178,000, April +148,000, May +129,000, June +57,000. Read the last four in order — that’s not a whipsaw anymore, it’s a slide. And it’s happening while companies redirect payroll into AI infrastructure. Wage growth near a five-year low is the tell.
Why this matters even if you’re employed: When wages stagnate while productivity rises, the gains flow to shareholders and executives, not workers. When millions of people have less money to spend, the businesses you work for have fewer customers. The restaurant near your office closes. Your company’s revenue drops. Your “safe” job becomes the next round of layoffs. Mass unemployment doesn’t stay contained - it spreads through the economy like a contagion.
Source: Bureau of Labor Statistics (June 2026), CNBC: June 2026 Jobs Report
The Counter-Signal: Job Postings Just Flipped
Here is the strongest available argument against everything above, and it deserves to be stated at full strength rather than waved past.
On July 8, 2026, Indeed’s Hiring Lab published a finding that reverses the direction of its own four-year trend. From May 2022 to May 2026, the rule was simple: the more exposed an occupation was to AI, the harder its job postings fell. Between May 2025 and May 2026, that relationship inverted. The more AI-exposed the occupation, the more it rebounded.
Software development, the sector this article has treated as the canary, is the clearest case. US software development postings are up roughly 15% since Claude Code launched in late February 2025, while overall job postings fell 7% over the same stretch. The pattern shows up across most developed economies, with English-speaking countries moving earliest — consistent with agentic-AI adoption driving it rather than coincidence.
If AI were simply eating programming jobs, this could not happen.
So what is happening? Look at the composition. 71% of the net increase is senior roles, and 37% is jobs with AI in the title, two categories that heavily overlap. And postings remain about 27.5% below pre-pandemic levels while overall postings sit roughly flat versus February 2020. Elsewhere in the data, workers aged 22-25 saw employment fall nearly 20% from late 2022 to mid-2025.
That is not a labor market recovering. It is a labor market recomposing — hiring back at the top of the ladder while the bottom rungs stay sawn off. Demand for people who can direct AI is real and rising. Demand for people who would have become those people by doing three years of junior work is not.
Which is worse, in a way the headline number hides. A market that destroys jobs uniformly is a shock you can plan around. A market that keeps the senior tier and eliminates the apprenticeship is a demographic time bomb: it works fine for a decade, then discovers it has no one qualified to promote. The cliff didn’t cancel. It moved to the entry gate. That dynamic — cuts driven by expectation of AI rather than deployed AI — is examined in anticipatory displacement.
The AI Acceleration: This Isn’t Your Father’s Automation
Two things make this different from every previous economic disruption: speed and scope.
The AI-Generated Code Reality Check
Software development is the canary in the cognitive coal mine. If AI can automate programming, the thing we told everyone to learn because “robots can’t code,” what’s safe?
| Metric | Figure | Source |
|---|---|---|
| Share of GitHub-committed code AI-generated/assisted | 51% (early 2026) | JetBrains research, industry surveys |
| Java code written by Copilot | 61% | GitHub |
| Developers using AI coding tools | 74% globally | JetBrains (Jan 2026) |
| Code suggestions kept in final submissions | 88% | GitHub |
| Task completion speed increase | 55% faster | GitHub Research |
| #1 AI coding tool by usage | Claude Code | (overtook Copilot in 8 months) |
| Fortune 100 Copilot adoption | 90% | GitHub |
Read that again: by early 2026, a majority of code on GitHub carries AI fingerprints. Three-quarters of working developers use AI tools daily. Claude Code went from launch (May 2025) to #1 most-used AI coding tool in eight months. Developers accept 88% of AI suggestions and complete tasks 55% faster.
Call it “assistance” if you want. It is replacement happening in slow motion, except it isn’t slow, only visible enough that we haven’t panicked yet.
Source: GitHub Blog, GitHub Blog: Economic Impact of AI-Powered Developer Lifecycle
McKinsey’s 30% Warning
McKinsey Global Institute, not exactly a fringe operation, projects that by 2030:
| Scenario | Hours Automated | Workers Needing Career Changes |
|---|---|---|
| Without generative AI | 21.5% | ~8 million |
| With generative AI | 29.5% | 12 million |
That’s an 8 percentage-point acceleration just from generative AI. Eight percent of all work hours in America, gone. Not “transformed,” gone.
“But won’t new jobs replace the old ones?” This is the “lump of labor fallacy” defense, the idea that there’s always a fixed amount of work, so automation just shifts it around. Historically, this was true: tractors displaced farm workers, but factories hired them; computers displaced typists, but created IT departments. The difference now is speed and scope. Previous transitions took 40-60 years, long enough for new industries to emerge and workers to retrain. AI is automating cognitive work across all industries simultaneously, in a decade. There’s no ladder to climb because the ladder itself is being automated.
The 12 million workers who need to switch careers by 2030 is 25% more than McKinsey projected just two years ago. The biggest shift falls on STEM professionals, the people we told to “learn to code,” who face automation potential jumping from 14% to 30% of work hours.
The advice to “get a tech job” is now about as useful as “get a factory job” was in 1975. No cohort feels this more acutely than new graduates: the 2026 class fears AI will take their jobs before they even land one, as entry-level rungs vanish.
Source: McKinsey Global Institute, Fortune
Goldman Sachs: 300 Million Jobs Exposed
Goldman Sachs economists estimate that 300 million full-time jobs globally could be exposed to AI automation:
- Two-thirds of jobs in the US and Europe face some AI exposure
- 7% of jobs could be entirely replaced
- 63% of jobs will be “complemented” (read: transformed beyond recognition)
- 30% of jobs remain unaffected (for now)
The optimistic spin: AI could increase global GDP by 7% over the next decade. The pessimistic reality: that GDP will be concentrated among those who own the AI systems, not those displaced by them.
What “concentrated” means for you: GDP measures total economic output, but says nothing about who receives it. If AI doubles economic output while eliminating half of jobs, GDP rises, but half the population has no income to participate in that “growth.” They become spectators to prosperity they can see but not access. This has already happened in sectors like finance: Wall Street profits hit records while Main Street wages stagnated for decades.
Source: Goldman Sachs
The DOGE Effect: Government as Preview
Remember when people joked about making government “run like a business”? Well, the Department of Government Efficiency (DOGE) is delivering on that joke, and the punchline is 293,753 federal and contractor jobs.
DOGE-related cuts are now the #1 cited reason for layoffs in 2025. Federal contractors are preemptively reducing headcount before contracts get cancelled. Non-profits dependent on government funding are shuttering programs. The ripple effects are hitting sectors that weren’t even on the automation radar.
March 2025 alone saw 275,240 announced job cuts, with 216,670 directly attributed to DOGE actions. That’s ideology with a spreadsheet rather than automation. But it’s creating the same outcome: millions of workers discovering that their jobs were more expendable than they thought.
Source: Challenger, Gray & Christmas
Why This Time Really Is Different
Every time someone cries “automation apocalypse,” skeptics point to history: the Luddites were wrong, the Industrial Revolution created more jobs than it destroyed, ATMs didn’t eliminate bank tellers. Why should AI be different?
The historical “markets adjust” argument in plain English: Economists observe that every past automation wave (looms, tractors, assembly lines, computers) initially displaced workers but eventually created more jobs than it destroyed. Factories replaced farms but hired more people. Computers replaced typists but created IT departments. The pattern was so reliable that “technology creates more jobs than it destroys” became economic orthodoxy.
But that pattern required time and somewhere to go.
Three reasons this time breaks the pattern:
1. Speed
| Era | Disruption | Adaptation Time |
|---|---|---|
| Industrial Revolution | Mechanization | ~60 years |
| Electrification | Factory automation | ~40 years |
| Computing | Digital transformation | ~30 years |
| AI Era (2020s) | Cognitive automation | ~10 years |
Previous disruptions gave humans generations to adapt. Factories didn’t appear overnight; children grew up knowing their skills would be obsolete. AI capability doubles roughly annually. Humans don’t evolve that fast.
2. Target
Every previous disruption attacked manual labor first, giving cognitive workers time to climb the skill ladder. AI is eating cognitive labor first. The lawyers, accountants, programmers, and analysts are getting hit before the plumbers and electricians.
There is no ladder to climb here. The ladder itself is being pulled up.
3. Scope
The Luddites destroyed textile looms. AI writes code, diagnoses diseases, generates marketing copy, handles customer service, trades stocks, drafts legal documents, and composes music. That reaches every industry that involves information work, not just one.
4. The Skills Gap Is a Chasm
New AI-related jobs require credentials most displaced workers don’t have:
| Requirement | % of New AI Jobs | % of US Adults |
|---|---|---|
| Master’s degree | 77% | 13% |
| Doctoral degree | 18% | 4% |
| Bachelor’s or less | 5% | 83% |
We’re creating jobs that 95% of displaced workers cannot fill. “Just retrain” might as well be “just become a different person” - and the track record of transition programs, from Finland’s cancelled UBI experiment onward, shows why retraining at this scale keeps failing.
The Vision: What Happens Next?
Elon Musk’s Bet
The tech leaders aren’t hiding what’s coming. Elon Musk has repeatedly predicted:
- “Universal High Income” (not just basic income) will become necessary
- Work will become optional within 10-20 years
- Money itself may become “irrelevant” as AI creates abundance
- He gives this an 80% probability
This goes past poverty prevention into a complete reimagining of economics. Musk’s philosophical challenge: “If the computer and robots can do everything better than you, does your life have meaning?”
His answer: work becomes voluntary, like “playing sports or a video game.” You do it because you want to, not because you’ll starve if you don’t. We run the actual math behind Universal High Income, and weigh the rival proposal of allocating compute instead of cash, in companion notes.
Source: Fortune
The Unscarcity Framework
In the Unscarcity framework, the Labor Cliff is the problem statement for the next civilization, not a catastrophe to be averted.
More job training programs won’t cut it (though they help at the margins). The real task is recognizing that an economy built on human labor is becoming obsolete, and building new systems that decouple survival from employment.
That’s what the Abundant Foundation and Impact accomplish: a two-tier system where everyone gets to exist (Tier 1 Residency through the Foundation), and those who want to contribute earn influence through Impact, a decaying currency that prevents oligarchy while still giving humans mountains to climb.
The choice we face isn’t “jobs vs. no jobs.” It’s “Star Wars” (elite capture of abundance technology) vs. “Star Trek” (abundance as infrastructure for all).
The 2025-2030 Labor Cliff is the gap between where we are and where we need to be. The question is whether we’ll build bridges or fall into the chasm.
What This Means for You
Immediate (2025-2026)
- Assume your job will change beyond recognition. Not “might” - will. Even if you’re not fired, your role in 2027 will look nothing like your role today.
- Learn to direct AI, not compete with it. The skill isn’t “doing the task better than AI.” It’s “knowing which tasks to give AI and how to verify the output.” This is the heart of agentic AI orchestration, the career-defining skill of 2026.
- Document your unique value. Complex emotional intelligence, creative problem-solving across domains, ethical judgment: these are still human advantages. For now.
Medium-term (2026-2028)
- Diversify income streams. One employer is a single point of failure in a volatile labor market.
- Build human networks. AI can’t yet replace trust relationships. Your network is your safety net.
- Consider geographic flexibility. Some regions will adapt faster than others.
Long-term (2028-2030)
- Redefine purpose. If work becomes optional, what gives your life meaning? Start answering that question now.
- Advocate for transition infrastructure. Support discussions about abundance frameworks, retraining programs, and economic transformation.
- Create rather than compete. Focus on uniquely human endeavors: art, connection, stewardship.
The Uncomfortable Truth
The companies laying off thousands while posting record profits aren’t being cruel. They’re adapting to a new reality where human cognitive labor is increasingly obsolete. You can be angry about it - anger is appropriate - but anger won’t change the thermodynamics.
What will change things is building new systems before the old ones collapse.
The 2025-2030 period is the Labor Cliff, the moment human work as we’ve known it begins its permanent transformation. Those who recognize this and adapt will navigate the transition. Those who don’t will become statistics.
The numbers are screaming. Are you listening?
Sources and References
Labor Market Reports
- Bureau of Labor Statistics Employment Situation (June 2026)
- Bureau of Labor Statistics Employment Situation (May 2026)
- Bureau of Labor Statistics Employment Situation (April 2026)
- CNBC: June 2026 Jobs Report - Payrolls Grow Just 57,000
- Indeed Hiring Lab: AI and Job Postings - From Destruction to Creation? (July 8, 2026)
- Challenger Report: March 2026 - AI Leads Reasons for First Time
- Challenger Report: April 2026 - Cuts Rise 38% from March, AI 26% of Total
- Challenger Report: May 2026 - Cuts Rise 16%, Highest May Since 2020
- Challenger Report: June 2026 - Layoffs Cool to 45,849, AI Leads for a Fourth Straight Month
- HR Dive: Tech Layoffs Surge 83% in H1 2026 as AI Restructuring Tops Challenger Report
- SiliconAngle: Meta Shifts 7,000 Employees into Four New AI Units
- CNBC: March 2026 Jobs Report
- CNBC: Meta + Microsoft 20K Cuts Raise AI Labor Crisis Concern
- Variety: Meta Layoffs 8,000 Employees
- Tom’s Hardware: Q1 2026 Tech Layoffs Hit 80K, ~50% Due to AI
- BusinessToday: April 2026 Tech Layoffs ~40K
- CNBC: Layoffs Top 1.1 Million in 2025
Tech Layoff Tracking
- TechCrunch: 2025 Tech Layoffs List
- Crunchbase: Tech Layoffs Tracker
- TrueUp: Layoffs Tracker
- Fortune: How Microsoft, Google, and Meta Are Plotting for the AI Era
- Fast Company: 1 Million Layoffs
AI and Automation Research
- McKinsey Global Institute: Generative AI and the Future of Work in America
- Fortune: McKinsey Projects 12 Million Job Switches
- Goldman Sachs: AI Could Raise Global GDP by 7%
- GitHub Blog: Does Copilot Improve Code Quality?
- GitHub Blog: Copilot for Business Is Now Available
- GitHub Blog: Economic Impact of AI-Powered Developer Lifecycle
Universal High Income and Future Visions
DOGE and Government Impact
Related Research
- What to Do Before the Labor Cliff: Action Guide for 2025-2030 - A concrete checklist for navigating the displacement covered above
- Maria’s Story: A House Cleaner Replaced by Robots - What the layoff statistics look like for one displaced worker
- How Large Language Models Work - The transformer technology driving cognitive-work automation
Related Articles
- Employment Statistics - Full statistical breakdown
- Musk’s Universal High Income - The case for abundance beyond UBI
- AI Coding Revolution - How AI is transforming software development
- The Humanoid Robot Revolution - The body of the revolution
- Robots That Build Abundance - Why the construction trades were never the safe side of the cliff, just the next one
- The Liability Gap: What a License Really Sells - Why licensed professions look immune right up until the accountability layer gets rebuilt, then go all at once
- The EXIT Protocol - How elites can land softly
- The Bootstrap Paradox - How to fund the transition before the window closes
Last updated: August 1, 2026 - added the June Challenger report (total cuts fell 53% to 45,849 while AI’s share rose to 31%, leading for a fourth consecutive month; 101,743 AI cuts and 139,156 tech cuts through H1), June BLS data (4.2% unemployment on 720,000 labor-force exits, +57K payrolls, April and May both revised down), and Indeed Hiring Lab’s July finding that the AI-exposure-to-postings relationship has flipped from destruction to creation - concentrated almost entirely in senior roles.
The cliff doesn’t wait for stragglers.
