AI and jobs · 2025

The career ladder is shortening at the bottom. The top is fine for now.

○◔◑◕●

Keep an eye on it

The disruption is real, particularly at entry level. Not an emergency today, but probably worth thinking about where you sit.

In a nutshell: A US bank and a European telco both announced headcount reductions this week and cited AI adoption explicitly. Those announcements are real. The leap to systemic crisis is not supported by the data yet.

Spin Unspin

What they’re saying

The comparison to the industrial revolution is structurally valid. The risk is that it misleads about pace and target: previous transitions took decades and hit physical labour. This one may be faster and is targeting cognitive work.

The substantive argument. AI is most disruptive to structured cognitive tasks — legal research, junior analysis, customer service — exactly the entry-level roles that middle-class career paths are built on.

Published figures range from 10% to 85% depending on methodology. That spread reflects genuine modelling uncertainty. Treat as scenario planning, not forecast.

This framing has appeared in three distinct waves since late 2022. Each wave is triggered by real announcements, amplified into systemic narrative, then quiets when the data doesn’t follow.

Hard data is limited — most figures come from modelling with wide confidence intervals. Individual company announcements are verified; sector-wide projections are not.

Do I care?

Your risk horizon is 5–10 years. Long enough to adapt deliberately. The disruption is real at entry level, not where you are.

Research, drafting, basic analysis — increasingly automatable. Worth being deliberate about whether you’re hiring for tasks or for the development of judgement.

Build skills complementary to AI tools rather than substitutable by them. Judgement, relationships, complex problem framing — these are the near-term hedges.

Otherwise, take it as an opportunity.

Was this relevant?

Noted