The catalogue
Every profile, with the evidence attached.
Each profile is a folder: a typed data file with the numbers and their sources, plus a written note with the argument. Where a role has no official projection, the card says so rather than inventing a figure.
24 shown
Home health and personal care aides
The largest source of new jobs in the US economy — and the one AI helps least.
Nurse practitioners
The fastest-growing occupation in the BLS table, at 41% — more than ten times the economy-wide rate.
Solar photovoltaic installers
Physical work in a place that changes every day, on a roof that does not read specifications.
Wind turbine service technicians
Eighty metres up a tower, on a machine that is not coming to you.
Industrial machinery mechanics
When the automation breaks, someone has to take it apart — which is a growing job.
Firefighters
Ranked the single most AI-proof occupation in a study of O*NET physical abilities.
Radiologists
The profession everyone predicted would vanish first, which instead got better at its hard half.
Teachers
Real-time reading of other humans, in a room, with a duty of care.
Data entry keyers
Fully specified input, verifiable output, no judgement — the pure case of automatable work.
Bookkeeping, accounting and auditing clerks
Rule-based verification of numbers, at high volume, in a digitised system.
Customer service representatives
The clearest case of AI eating the entry rung rather than the ladder.
Interpreters and translators
Ranked the single most AI-affected occupation in Microsoft's usage data.
Software developers
Split in two: the ones who direct agents are thriving; the ones who wrote boilerplate are not.
Administrative assistants and executive assistants
Scheduling, drafting and coordinating — the exact activities appearing most in real AI usage.
Junior data analysts
The role that proves the entry level is being redesigned rather than deleted.
AI engineer (applied / LLM)
The single fastest-growing job title in the US, two years running.
Forward deployed engineer
Postings up 729% in a year: the job that exists because AI projects fail for organisational reasons.
Data annotators and AI trainers
The most accessible on-ramp into AI work, and the one that rewards domain expertise over coding.
AI governance, risk and audit
Regulation creates a function; the function creates the jobs.
Agent fleet manager
No one holds this title yet. The pattern it would run — one coordinator delegating work to a fleet of subagents — is already shipping in software.
Robot fleet supervisor
Every robot that arrives creates people who watch, fix and redirect it. The fleet-scale version of that shift has no title yet.
AI auditor
The build half of AI hiring runs about seven to one ahead of the governance half. Someone has to close that gap, and the role has no settled name.
AI output verifier
The quiet rate-limiter on every agent deployment: a person who can tell when the machine is wrong, at volume, in their own domain.
Agent liability underwriter
When an agent books the slot or signs the order, someone has to price the mistake. No market for that exists yet.
Nothing here is a prediction of a specific person's employment. The profiles describe task profiles and the demand for the role — which is a different, and much more defensible, claim.