Molly Kinder, former Brookings senior fellow, warns that AI‑driven automation will first hit high‑paid white‑collar jobs, creating a politically volatile “messy middle” before any post‑AGI abundance. She proposes targeted interventions – workforce reinvestment funds, wage insurance, and a new industrial policy for knowledge work – as alternatives to universal basic income.
Business news
Molly Kinder, who led Brookings’ multi‑year study of generative AI’s impact on work, announced she is leaving the think‑tank to launch an organization focused on solving the AI transition. In a recent essay she coined the term “messy middle” to describe the period between today’s largely intact labor market (Reality 1) and a speculative post‑AGI abundance (Reality 3). In this intermediate phase (Reality 2) most jobs survive, but high‑paid, computer‑based roles – law, finance, consulting, software engineering – face concentrated displacement.
Kinder argues that this concentration will be “politically explosive” because the very workers who have traditionally benefitted from automation (the “laptop class”) will be the first to lose their livelihoods. She rejects universal basic income (UBI) as a blunt tool that would “destroy the labor market” by paying displaced knowledge workers more than essential workers, and instead calls for targeted, demand‑side policies.
Market context
- Task‑exposure data released by OpenAI shows that roughly 45 % of tasks in professional services (legal research, financial modeling, market analysis) are highly automatable with large language models, compared with less than 10 % in manufacturing or hospitality.
- The U.S. labor market still leans heavily on these knowledge occupations: the Bureau of Labor Statistics reports ~30 % of all private‑sector jobs require a bachelor’s degree, and median earnings for these roles sit at $92,000 (2023).
- The U.S. safety net remains thin: unemployment insurance replaces only 40 % of prior earnings on average, and the average duration of benefits is 12 weeks.
- Historical parallels – the de‑industrialization of the 1980s – show that concentrated job loss in a single sector can trigger long‑lasting political backlash, as seen in the rise of protectionist sentiment in the Rust Belt.
What it means
1. Policy levers that slow the pace of displacement
- Workforce reinvestment fund – Companies that cut entry‑level knowledge positions would be required to contribute to a fund that finances white‑collar apprenticeships. Similar schemes exist in Michigan, where Rep. Mallory McMorrow’s legislation ties corporate tax credits to apprenticeship creation.
- Retention‑first regulations – Mandating that firms document “re‑skilling pathways” before layoffs could raise the cost of rapid headcount reductions, giving workers a buffer period to transition.
2. Safety‑net upgrades for displaced professionals
- Wage‑insurance programs – Modeled on Germany’s “Kurzarbeit” scheme, wage insurance would top up a worker’s earnings when they take a lower‑pay job after displacement, preserving household consumption and mortgage solvency.
- Extended UI benefits – Expanding the duration and replacement rate for high‑skill workers (e.g., from 12 weeks to 26 weeks, and from 40 % to 60 % of previous earnings) would mitigate the steep income shock seen in case studies like the former USAID official who faced a 60 % pay cut to become a teacher.
3. A new industrial policy for knowledge work
- Publicly funded “social impact labs” – Government grants could seed teams that apply AI to solve chronic societal problems (e.g., affordable housing analytics, climate‑risk modeling) while paying market‑rate salaries. This mirrors the New Deal‑era Works Progress Administration but for the digital economy.
- Capturing AI surplus – Taxing a modest share of AI‑generated productivity gains and directing the revenue to the reinvestment fund would create a sustainable financing loop.
4. Why UBI falls short in the messy middle
Kinder notes that a universal check large enough to replace a senior software engineer’s salary would disincentivize essential low‑wage work and could erode the labor market’s price signals. Instead, means‑tested, role‑specific support preserves the incentive structure while addressing the acute pain points of displaced knowledge workers.
Strategic implications for businesses
- Talent‑risk assessments should now include AI‑exposure scores for each role, similar to the OpenAI task‑exposure matrix, to anticipate where turnover risk spikes.
- Companies that invest early in upskilling and apprenticeship pipelines may gain a competitive edge by retaining institutional knowledge and avoiding costly recruitment cycles.
- Firms that embrace responsible AI guardrails—as Anthropic does with Claude Fable 5—can differentiate themselves to talent wary of abrupt displacement.
Looking ahead
The “messy middle” could span several decades, depending on the speed of model improvements. While the full‑scale job apocalypse remains speculative, the near‑term concentration of pain in high‑skill, computer‑based occupations is already observable. Policymakers, employers, and educators must therefore act now to shape a transition that preserves the American middle class rather than erodes it.
Further reading
- Kinder’s original essay on the “messy middle” – Brookings Report
- OpenAI task‑exposure dataset – OpenAI Blog
- Anthropic’s guard‑rail approach for Claude Fable 5 – Anthropic Announcement
- German Kurzarbeit model – Federal Ministry of Labour
Casey Newton’s Platformer podcast interview with Molly Kinder is available on all major platforms. Feedback can be sent to [email protected].
Comments
Please log in or register to join the discussion