Latest / Minds, Bodies, and Terawatts / California Becomes First State to Launch Real-Time AI Job Loss Tracker
Transcript
- 0:00[Host] This is Minds, Bodies, and Terawatts, where the state that builds AI just started counting who it replaces. Bloomberg reported on June 25th that California became the first state to launch a real-time tracker connecting AI exposure to actual unemployment claims.
- 0:16[Guest] Governor Newsom partnered with the Employment Development Department and the California Policy Lab, a UC research institute. The tool maps monthly unemployment claims against how much AI exposure each occupation carries, and the dashboard is publicly accessible.
- 0:30[Host] How is this different from the layoff reports we already get?
- 0:34[Guest] Challenger and BLS count losses after they happen. This tracker correlates exposure risk with claims data as it comes in, asking whether filings are rising faster in high-AI-overlap occupations than in low-overlap ones. It is an early-warning instrument, not a postmortem.
- 0:51[Host] What does the early data actually show?
- 0:53[Guest] Ben Hyman, senior researcher at the California Policy Lab, said there is no evidence of large-scale AI layoffs statewide. But college-educated workers in high-exposure roles, specifically in the Bay Area and tech-heavy sectors like information services, are filing claims at elevated rates. The signal is localized.
- 1:11[Host] The first population showing stress is college-educated tech workers, not warehouse staff?
- 1:16[Guest] AI replaces decisions, drafts, and code before it replaces physical tasks. The workers whose daily routines overlap most with what AI can do today carry degrees and salaries. That matches every major layoff report from this year.
- 1:31[Host] Newsom preceded the tracker with a May 21st executive order directing state agencies to prepare for AI workforce disruption. In the announcement, he said California is shaping the future and charting the course for the nation.
- 1:45[Guest] He is also weighing a 2028 presidential run, and his state hosts OpenAI and Anthropic. The governor whose state builds the technology is now the first governor tracking its labor impact. Whether the motive is policy or politics, other states can copy the tool because it runs on existing unemployment insurance data.
- 2:05[Host] What can't the tracker detect?
- 2:07[Guest] Workers who get reassigned internally or who quit because their roles shrank. If your employer moves you from writing reports to managing the AI that writes them, no claim gets filed. That quiet substitution is invisible to the dashboard.
- 2:22[Host] Picture a data analyst who used to build financial models all day. Now she proofreads what the AI drafted. Her role shrank by half, but she never filed a claim.
- 2:34[Guest] The tracker measures correlation, not causation. But as an early-warning system, it is the first government instrument asking the question occupation by occupation, in near real time.
- 2:45[Host] The mechanism underneath is simple. Once the cheapest reliable provider of a task stops being human, that job crosses a line it never crosses back. And it does not happen economy-wide on one date.
- 2:58[Guest] It happens one help desk, one coding team, one service queue at a time. The tracker is built to catch those individual crossings accumulating before they reach the national unemployment number.
- 3:10[Host] And California's early data puts the crossings in the occupations everyone assumed were automation-proof.
- 3:16[Guest] The state that builds the AI is the first state measuring who it displaces, and the early answer landed where almost nobody expected, not on factory floors but in offices with diplomas on the wall.
- 3:30[Host] Patrick Deglon calls the accumulation of those crossings the Labor Cliff in Unscarcity.
- 3:35[Guest] For more on what happens when those crossings reach critical mass, head to unscarcity dot ai. The book is on Amazon, the audiobook is on Spotify, and this has been Minds, Bodies, and Terawatts.