Two Futures. One Decade. Your Choice.

Anthropic just published the most rigorous labour market study yet of what AI is actually doing to professional work. Not what it could theoretically do. What it is doing right now, in real workplaces, tracked across millions of interactions.

The headline finding sounds reassuring. No mass unemployment. The labour market has not broken.

Do not be comforted by that.

Hiring of workers aged 22 to 25 in AI-exposed occupations has fallen 14 per cent since ChatGPT launched. The entry-level roles that once trained the next generation of lawyers, analysts and developers are quietly disappearing. The workers most exposed are not low-skilled. They hold graduate degrees. They earn above the median. They are disproportionately women.

In this episode David Richards MBE examines two very different futures. The bull case: AI drives genuine abundance, intelligence becomes democratised, and the productivity gains are large enough to fund a serious universal basic income. The bear case: firms capture the gains as margin, growth continues but prosperity decouples from it, and an entire generation gets sorted into winners and losers before anyone in power notices.

The China trade shock showed up in employment data in 2003. The political reckoning arrived in 2016. Thirteen years between the signal and the crisis.

The AI signal started in 2022.

The difference between the two futures is not the technology. The technology arrives either way. The difference is the choices being made right now.

The Sunday Signal is published every week at newsletter.djr.ai. Subscribe for free to get every issue directly to your inbox.