Block cut 4,000 workers, then quietly brought some back with raises. Klarna replaced 700 with AI, then admitted it "went too far." The layoff wave is real — and so is the rehiring.

Block CEO Jack Dorsey cut roughly 4,000 employees in early March 2026 (about 40% of the company's workforce) and told analysts AI models had gotten "an order of magnitude more capable." Block's stock surged as much as 24% on the news. Two weeks later, it was down 6%, as the market priced in execution risk. Some laid-off workers were quietly brought back. At least one remaining employee received a retention package increasing pay by tens of thousands of dollars. The layoff was real. The rehiring was real too. And the gap between those two facts is where the actual story of 2026's labor market lives.

A Skills Crisis, Not a Jobs Crisis

Over 92,000 tech workers have lost their jobs in 2026 so far, pushing the total to nearly 900,000 since 2020, per Layoffs.fyi data reported by CNBC. But the dominant narrative (that AI is simply eliminating work) is contradicted by what companies are actually doing. Gartner predicts that by 2027, half of the companies that attributed headcount reductions to AI will rehire for similar functions, often under different titles. A Robert Half survey of 2,000 hiring managers found that 29% have already reopened positions they eliminated after implementing AI. Analytics firm Visier tracks a rehire rate of 5.3% for terminated employees, and it's "already ticking up."

The real story is not jobs disappearing. It is jobs migrating — into hardware, robotics, and physical-AI companies that demand hybrid skill sets most displaced pure-software engineers don't yet have. The professionals who understand this shift will land the highest-paying roles in the next 18 months.

Yes, the Layoffs Are Real — but AI Isn't the Whole Explanation

The scale of the 2026 cuts is historically large. U.S. employers announced over 61,650 job cuts in the last week of January alone — the highest January level since 2009, per Challenger, Gray & Christmas data reported by Fast Company. AI has been cited in 71,825 job cut announcements since 2023, HCAMag reports. Meta is cutting about 8,000 jobs and scrapping 6,000 open roles. Microsoft confirmed voluntary buyouts for roughly 8,750 U.S. employees, its first such program in 51 years. Snap is slashing 1,000 staffers. Salesforce cut 4,000 customer support roles. Atlassian cut 1,600 workers. Pinterest cut almost 15%. Oracle eliminated an estimated 20,000 to 30,000 positions. Amazon has cut at least 30,000 jobs since October 2025. Nike cut 775 distribution center workers. Home Depot laid off 800 employees.

But the executives making these cuts don't all believe their own press. Sam Altman, speaking at BlackRock's US Infrastructure Summit, called the widespread AI blame "AI washing," per CEOWORLD magazine. Wharton professor Peter Cappelli said, "We spend a lot of time looking carefully at companies that are actually trying to implement AI, and there's very little evidence that it cuts jobs anywhere near like the level that we're talking about." Marc Andreessen put it more bluntly on a podcast: large tech companies were overstaffed, and "now they all have the silver-bullet excuse: ah, it's AI," as reported by The Guardian.

Amazon's SVP Beth Galetti told CNN that AI isn't the reason behind the "vast majority" of the company's 30,000-plus cuts, framing them as "reducing layers, increasing ownership, and removing bureaucracy." A laid-off Amazon employee was more direct in an interview with ABC News: "You could potentially have just been bloated in the first place, reduce head count, attribute it to AI, and now you've got a value story."

The research backs up the skepticism. MIT found that 95% of AI pilot programs had not led to increased productivity or savings, The Guardian reports. UC Berkeley research shows AI is intensifying work rather than reducing the need for human labor. A Carnegie Mellon study found that even the best AI workers could only complete about a quarter of basic tasks assigned to them, Forbes reports.

The layoff wave is real. The claim that AI is the primary cause is, at best, incomplete.

Klarna Went Too Far — Then Backtracked

Klarna offers the clearest case study. Between 2022 and 2024, the fintech company eliminated roughly 700 customer service positions, replacing them with an AI assistant built with OpenAI. By May 2025, CEO Sebastian Siemiatkowski admitted the company had "gone too far." His explanation was blunt: "We focused too much on efficiency and cost. The result was lower quality, and that's not sustainable." Klarna reversed course and began rehiring, targeting students and rural workers for remote roles. The company is now building what Siemiatkowski called "an Uber type of setup," recruiting customers for gig customer-service roles to handle questions the AI can't.

Klarna is not an outlier. Gartner's survey of 321 customer service leaders found only 20% had actually reduced staffing because of AI, despite widespread claims, Forbes reports. Robert Half found 55% of hiring managers planned to increase contract or temporary workers in the first half of 2026.

At Block, the rehiring was quieter but telling. A handful of laid-off workers returned after the 4,000-person cut, Business Insider reports. At least one remaining employee got a retention package increasing pay by tens of thousands — a signal that the company knew it needed to hold onto specific people even as it shed headcount.

The rehiring, however, is not happening on the same terms. A former Microsoft worker was laid off, then watched his team get restaffed with contract workers. He eventually returned in a lower-ranked role paying about a third less. "My morale has taken an enormous hit," he told Business Insider.

The layoff-rehire cycle is already underway. The critical wrinkle is that the jobs coming back look different from the jobs that were cut.

Where the New Jobs Pay Six Figures

The jobs being created are not in the same places the jobs were lost. They are migrating into robotics, autonomous systems, and physical-AI infrastructure — and they pay significantly more, but require hybrid skill sets that pure-software engineers typically lack.

OpenAI is actively hiring robotics engineers to build robots that can help construct data centers, power grids, and other critical infrastructure. Sam Altman posted on X on May 31, 2026, seeking "exceptional full-stack hardware, ops, systems, and ML engineers." Zero G Talent's job board lists 50 OpenAI roles added in the past week, including an MEP Coordinator, Mechanical - Stargate (Remote - US) at $164,000–$268,000 per year and an AI Tooling & Enablement Manager in San Francisco at $207,000–$230,000. Anthropic added 40 roles in the same period, including a Product Manager, Enterprise at $385,000–$595,000 and an Engineering Manager, Safeguards Review Tooling at $405,000–$485,000.

The robotics salary data makes the opportunity concrete. A CareersInRobotics.com analysis of 907 jobs found the median salary in robotics is $156,563. Software-focused roles average $194,000. Hardware-focused roles average $127,000 — a 53% premium for software skills applied to physical systems. The highest-paying skills in robotics tell their own story: CUDA carries a 45% premium, Technical Leadership 43%, Machine Learning 40%, C++ 37%, Python 37%. The highest-paying industry is Transportation & Autonomous Vehicles at a $200,000 median, followed by Robotics Software & AI at $198,000. California dominates compensation at $193,000 median — 81% above the non-California average of $106,000.

The investment is following the talent. Fanuc America announced a $90 million investment for a new 840,000-square-foot robot manufacturing facility in Michigan, targeted for completion in late 2027, Robotics and Automation News reports. The global stock of operational robots has doubled in the past five years, and robotics engineer job postings rose nearly 25% year-over-year in 2023.

Alphabet, Microsoft, Meta, and Amazon are expected to spend nearly $700 billion combined in 2026 on AI infrastructure buildouts, CNBC reports. Token costs for AI inference have dropped 280-fold over the past two years, The Guardian reports. The agentic AI market has grown to $9 billion, BCG reports.

The money is flowing. The jobs are being created. They are just not where the laid-off workers are looking.

Why Pure-Software Engineers Are Struggling to Pivot

The displaced engineers who will struggle most are not those who lack AI literacy. It's those who lack the physical-systems, hardware, and domain-specific knowledge that robotics and physical-AI roles require. The gap is structural, not cyclical.

The World Economic Forum's Future of Jobs Report 2025 finds that 39% of core skills are expected to change by 2027. Technological skills are projected to grow in importance more rapidly than any others, with AI and machine learning literacy, data analytics, and cybersecurity among the top hard skills. But creative thinking, resilience, flexibility, curiosity, and lifelong learning are among the top soft skills rising in importance.

BCG analysis projects that over the next two to three years, 50% to 55% of jobs in the U.S. will be reshaped by AI. Five years from now, 10% to 15% of jobs could be eliminated. McKinsey estimates nearly 12 million Americans will be forced to switch jobs by 2030 due to shrinking demand in their current occupations, with lower-income workers up to 14 times more likely to need a career change, Business Insider reports. The World Bank estimates roughly 1.1 billion workers will need retraining in the next decade.

Technology skills' half-life is now about 2.5 years, Deloitte reports. Princeton researcher Stephan Rabanser said reliability is the key limiting factor: "This is the barrier to job transformation," per The Guardian.

The gap is not about learning to use AI tools. It's about the fundamentally different knowledge domains (hardware, physical systems, safety-critical engineering, domain context) that pure-software engineers haven't been trained in.

The Contractor Economy: Work Restructured, Not Restored

Even when companies rehire, they are increasingly doing so through contract, gig, and contingent arrangements. This is a structural shift that gives employers flexibility but strips workers of stability, benefits, and bargaining power.

Upwork research found that 77% of business leaders say the AI era is increasing their need to hire contract workers with specialized skills, Business Insider reports. MBO Partners estimates 73 million people work as independents in the U.S., with contingent workers at roughly 40% of the workforce. Middle management job postings dropped more than 40% between April 2022 and October 2024, as companies create flatter structures, Deloitte reports. Thirty-six percent of managers expect to be managing digital agents within the next five years.

The former Microsoft worker who was laid off said a third-party contractor company reached out about working for the same team soon after. He passed on the opportunity. Contract workers often miss out on health insurance, 401(k)s, unemployment insurance, stock options, and stability. They have less recourse for harassment or discrimination.

David Weil, an economics professor at Brandeis University, said: "It's just part of this larger dance... where very profitable organizations want to share as little as they can with the people who create a lot of the value." Microsoft settled a class action lawsuit for $97 million in 2000 after contract workers argued they had been employed for too long to not receive benefits. As of 2019, Google had more temporary workers than full-time employees, The New York Times reports.

The rehiring is real. The form it takes is shifting beneath workers' feet.

Net Job Growth — but the Gains Are Concentrated

The aggregate data shows net job creation, not destruction. But the gains are concentrated in roles and geographies that displaced tech workers are poorly positioned to access without deliberate reskilling.

The World Economic Forum forecasts that by 2030, 92 million jobs will be displaced by AI and automation globally, while 170 million new jobs will be created — a net gain of 78 million, but with 22% of today's jobs fundamentally reshuffled, per a LinkedIn article citing WEF data. Goldman Sachs estimates roughly 300 million jobs globally could be affected by generative AI by 2030, covering about 18% of work hours worldwide. The IMF estimates 60% of jobs in advanced countries are exposed to AI-driven task automation, versus about 40% in emerging economies.

A Vanguard report found that jobs highly exposed to AI automation are actually growing faster than they did before the pandemic, and faster than all other occupations, Forbes reports. The WEF projects AI and Machine Learning Specialist roles will increase 40% by 2027, cybersecurity roles about 31%, and data analysts/scientists and digital transformation specialists about 30% each. The green transition and demographic shifts will create about 170 million new jobs by 2030.

The Bureau of Labor Statistics projects software developer employment will grow 15% by 2034, CNN reports. IBM is tripling entry-level hiring in the U.S. Intuit is hiring more early-career developers who have grown up using AI. Indeed listings for software engineer jobs are up 11% annually, though new postings declined 15% in the first two months of 2026 compared to the same period in 2025, per LinkedIn data reported by The Guardian.

But the transition is brutal. Forty percent of white-collar workers who changed jobs at the end of 2025 took pay cuts of 10% or more — the highest proportion in a decade, Business Insider reports. Glassdoor's Employee Confidence Index for tech fell 6.8 percentage points in March 2026 to 47.2%, the largest year-over-year drop of any industry.

The economy is creating more jobs than it's destroying. The "net gain" number is cold comfort to a displaced worker who lacks the skills for the new roles.

What the Winners Are Doing Right Now

The professionals who will capture the highest-paying positions in the next 18 months are those who deliberately build hybrid AI-physical skill sets now. A growing ecosystem of corporate, academic, and government programs is emerging to support this transition, but it's not moving fast enough.

IBM's general manager of automation and AI, Neel Sundaresan, said the job at IBM has shifted from routine coding to working directly with customers and specifying features that can be created with AI, CNN reports. Intuit CTO Alex Balazs said engineers now focus on more complex aspects of code and software design, spending more time solving customer problems. CoderPad CEO Amanda Richardson put it this way: "The job will look different. That doesn't mean it's going away. The best engineers are spending all day, every day with AI and using it to make their designs better."

Magdalena Balazinska, director of the University of Washington's Paul G. Allen School of Computer Science & Engineering, wrote to over 2,000 undergraduates: "AI is not killing your job options. It's expanding them." She added: "Roll up your sleeves... There will be many more technological breakthroughs from now until you retire."

Boston University's James Bessen offered a longer historical view: "New technologies don't just replace labor with machines — they also reduce prices and improve product quality. This increases customer demand and drives up employment."

The institutional response is real but uneven. WEF found that 85% of employers plan to accelerate upskilling or reskilling, per a LinkedIn article. IBM pledged to train 30 million people globally by 2030 in tech skills. Singapore's SkillsFuture program gives citizens training credits. The EU and various countries have announced national AI strategies with workforce development funds.

Venture capitalists report that companies can now reach $50 million in revenue with 50 employees, versus 250 people previously, CNBC reports. Meta's Reality Labs team was restructured into "AI-native pods" — smaller, more nimble, AI-first teams, Business Insider reports. Google credited AI for 50% of its code in its latest earnings report, The Guardian reports. Block's head of engineering said 90% of the company's code submissions were authored "partially or fully with AI support."

The tools are here. The companies are hiring. The question is whether the workforce can close the gap in time.

The Real Story Behind the Headlines

Block's stock surged 24% on the layoff news, then gave back most of the gain within two weeks. Oracle's stock popped 7.5% on its layoff announcement, then retreated to near pre-layoff levels within days, The Guardian reports. The market's verdict was consistent: a layoff announcement is not a strategy.

Zero G Talent tracks 11,513 open AI roles across 2,793 companies right now, and the fastest-growing category is the one that didn't exist five years ago: the hybrid role that lives at the intersection of software intelligence and physical systems. The median robotics salary is $156,563. Fanuc America is pouring $90 million into a Michigan manufacturing facility. The companies building the physical infrastructure of AI are hiring — not cutting.

The 18-month window is open. The question is not whether the jobs will exist. It is whether the people applying for them will have the skills to do them.

Working in AI? Zero G Talent tracks the openings: browse AI jobs, openings at OpenAI and Anthropic, and the people building the field.