In 1987, Nobel laureate economist Robert Solow looked at two decades of massive investment in personal computers and made a now-famous observation: "You can see the computer age everywhere but in the productivity statistics." Despite billions poured into IT throughout the 1970s and 1980s, the expected productivity boom simply wasn't showing up in the numbers. This became known as the Productivity Paradox (or Solow Paradox) - the persistent gap between technology investment and measurable productivity gains.

The paradox did eventually resolve. The U.S. saw a genuine productivity surge in the late 1990s and early 2000s, roughly 10-15 years after the PC became ubiquitous, once organizations stopped using computers as faster typewriters and started redesigning entire workflows around networked computing. And critically, while specific job categories were reshaped or eliminated during that transition, overall employment was not devastated - roles were replaced by others. Bank tellers gave way to ATM technicians and financial advisors. Typists became data analysts. The labor market absorbed the shock through role transformation, not mass unemployment.

Today, with generative AI, we appear to be living through the early chapters of the same story. The data - measured outcomes first, then the trajectory they point toward - suggests history is rhyming.

AI Adoption: Fast, But Not Yet Transformative

The Stanford 2025 AI Index Report found that 78% of organizations used AI in at least one business function in 2024, up from 55% the year before (Stanford HAI, 2025). The Federal Reserve Bank of St. Louis measured that 28% of U.S. workers personally used generative AI at work in the second half of 2024 (St. Louis Fed, 2025). Penn Wharton noted that these early adoption patterns are broadly similar to the adoption of personal computers in the early 1980s (Wharton, 2025).

If the PC analogy holds, we are roughly where the economy was in 1985. Adoption is visible everywhere. The productivity payoff is still years away.

Productivity: Individual Gains, Aggregate Silence

Workers who use generative AI report saving 5.4% of their work hours - about 2.2 hours per week. But when you factor in the 72% who don't use it, the aggregate saving across the entire U.S. workforce drops to just 1.4% (St. Louis Fed, 2025).

U.S. nonfarm labor productivity grew 2.3% in 2024 (BLS, 2025). Total factor productivity grew 1.3% - back to the long-term 1950-1999 average of 1.4%, and above the sluggish 0.9% post-2000 trend (BLS / CRS, 2025). Solid, but not the breakout everyone's waiting for.

Brookings reviewed the research and found that AI investments have not yet been associated with increases in sales per worker or revenue TFP. The main effect has been sales growth - firms capturing market share, not producing more with less (Brookings, 2025). PwC's analysis of close to a billion job ads found AI-exposed industries saw 27% growth in revenue per employee versus 9% in the least exposed (PwC, 2025) - but revenue per employee isn't the same as true productivity.

Classic Solow Paradox. Again.

Headcount: Reshuffling, Not Collapsing

MIT Sloan research using ~58 million LinkedIn profiles found that, as of late 2023, AI had not caused major changes in total employment. Firms with heavy AI use saw 6% higher employment growth and 9.5% more sales growth over five years (MIT Sloan, 2025).

But within firms, the mix is shifting. Top-paying analytical roles fell ~3.5%, while business and financial roles shrank 2-2.5% (MIT Sloan, 2025). U.S. customer service employment declined by ~80,000 positions between 2022 and 2024 (AIMMultiple, 2025).

The sharpest impact is on entry-level workers. Between late 2022 and mid-2025, entry-level employment in software engineering and customer service dropped roughly 20%, while employment for older workers in those same roles grew (CBS / Stanford, 2025). Globally, entry-level postings have fallen 29% since January 2024 (Randstad / WEF, 2025).

On the creation side, ITIF estimated AI directly generated ~119,900 U.S. jobs in 2024 - mostly in data center construction (ITIF, 2025).

Just like the PC era: specific roles contract, new ones emerge, and total employment absorbs the shock - but the transition is painful for those in the contracting categories.

Roles Fading, Roles Emerging

The WEF Future of Jobs Report 2025 names the fastest-declining roles: cashiers, administrative assistants, printing workers, accountants, and - new to the list - graphic designers. The fastest-growing: big data specialists, fintech engineers, AI/ML specialists, and software developers (WEF, 2025).

Entirely new categories are appearing at triple-digit rates. Autodesk's analysis of ~3 million listings found AI Engineer up 143.2%, Prompt Engineer up 135.8%, and AI Content Creator up 134.5% year over year (Autodesk, 2025). Veritone counted 35,445 AI-related U.S. positions in Q1 2025 (↑25.2% YoY), median salary $156,998 (Veritone, 2025).

PwC found skills in AI-exposed jobs are changing 66% faster than in less exposed roles. Workers with AI skills commanded a 56% wage premium in 2024, up from 25% in 2023 - while formal degree requirements for AI-augmented jobs fell from 66% to 59% between 2019 and 2024 (PwC, 2025). Harvard Business School found automation-prone roles saw 7% fewer skills in postings, while augmentation-prone roles gained new AI skill requirements (HBS, 2026).

The Scorecard: Fading vs. Emerging Roles at a Glance

The table below consolidates BLS occupational data (May 2024), job posting trends, and salary figures from the sources cited throughout this article. The Employment Trend column uses BLS 2024-2034 projections for fading roles and measured year-over-year posting growth for emerging roles. The Salary Trend column captures the directional movement based on BLS wage data and industry salary reports.

Sources: BLS Occupational Outlook Handbook & OEWS (May 2024), Veritone Q1 2025 Labor Market Analysis, Autodesk/GlobalData 2025 AI Jobs Report, PwC 2025 Global AI Jobs Barometer, Glassdoor (Dec 2025), Salary.com, Burtch Works (2025), WEF Future of Jobs Report 2025.

The pattern in this table is the same pattern the Solow Paradox describes in macro: the old roles are declining slowly and their wages are stagnating, while the new roles are growing explosively and their wages are rising sharply. The transition is underway. The question is speed - and whether individuals and organizations adapt fast enough to land on the right side of the table.

Can We Predict the Next Decade's Scorecard?

The data above is backward-looking. Now let's stress-test what happens if the pattern continues. This is explicitly a forward-looking exercise, and I'll be transparent about where each projection comes from. The Solow Paradox teaches us that the productivity payoff arrives roughly 10–15 years after adoption begins, and that the labor market absorbs the shock through role replacement rather than net job destruction. If that pattern holds, by 2035 we should be entering the productivity breakout phase - and the role reshuffling should be largely complete.

The WEF's Future of Jobs Report 2025 provides the macro frame: 170 million new jobs created globally by 2030, 92 million displaced, for a net gain of 78 million (WEF, 2025).

On the creation side, the WEF identifies several entirely new role families expected to scale by 2030: AI system architects, AI ethics and governance specialists, human-AI collaboration designers, and physical AI specialists working in robotics and autonomous mobility (WEF, 2025). A 2025 WEF governance report projects 5 million AI governance roles by 2030, driven by rising regulatory demand (the EU AI Act, similar frameworks in China and the U.S.) (WEF / WinsSolutions, 2025). McKinsey further expects a surge in demand for green economy roles, cybersecurity professionals, and healthcare workers augmented by AI (McKinsey, 2023).

Simultaneously, the roles currently showing early-stage decline are projected to accelerate their contraction. BLS projects that general-purpose IT support and routine software development will face disruption as AI autonomously generates, tests, and deploys code for standard tasks (WEF, 2025). Entry-level postings have already fallen 29% since 2024 (Randstad), and the IDC/Deel survey found 66% of global enterprises plan to cut entry-level hiring due to AI (IDC/Deel, 2025).

Here is a projected scorecard for ~2035, based on the trajectory of current data and the sources above:

Important caveat: This table is directional, not definitive. The Solow Paradox itself teaches us that the timing and magnitude of technology-driven shifts are notoriously hard to predict - economists in the 1980s were wrong about computers for over a decade. These projections combine BLS occupational outlooks, WEF employer surveys, and McKinsey task-automation estimates, but the actual pace depends on adoption curves, regulation, and organizational change that no model can fully capture.

The balance sheet tells the Solow Paradox story once more: the bottom of the table shrinks, the top of the table grows, and the net is positive - but only for those who are positioned on the right side of the transition.

The Paradox Will Resolve - It Always Does

The data is consistent: individual gains are real, role transformation is underway, aggregate productivity hasn't shifted yet. If the PC-era timeline holds, we have roughly five to eight years before the full reshuffling is complete. That window is not a reason to relax. It's a reason to move.

The scorecard makes the direction clear. The wage premium for AI skills doubled in a single year. Degree requirements are dropping while skill requirements are rising. The market is already telling you what it values next.

So if you're a developer who hasn't built anything with an AI API, start this week. If you're a designer who hasn't learned prompt-driven workflows, start today. If you're in a role the BLS projects to contract, those projections are not abstractions - they're a countdown. Explore adjacent roles. Learn to work alongside AI systems, not in competition with them. Treat 2025 as a training year, not a waiting room.

The Solow Paradox always resolves. History is equally clear on who it rewards: the people already in motion when the breakout hit - not the ones who saw it coming and stood still.