what workers is ai replacing?
the measurable displacement so far is in data entry, basic translation, first-draft copywriting, tier-one support, and junior document review. in most cases tasks were removed rather than whole roles, and the effect concentrates at entry level, where the work agents do well is exactly the work new people used to learn on.
where displacement is actually visible
| work | what changed | current state |
|---|---|---|
| data entry and extraction | vision models read documents reliably | largely automated, roles shrinking fast |
| basic translation | quality passed the threshold for routine text | translators moved to post-editing and specialised work |
| first-draft copywriting | volume content became near-free to produce | freelance rates for generic copy collapsed, editing held |
| tier-one support | deflection rates rose sharply | fewer tier-one seats, more escalation specialists |
| junior document review | pattern matching over contracts | the pyramid narrows at the bottom |
| stock photography and simple illustration | generation replaced licensing for filler | commissioned and brand work largely intact |
The pattern repeats: the commodity end of a market goes first, and the end that requires judgement, accountability, or a relationship holds.
the entry-level concentration
This is the effect that matters most and gets the least coverage. Almost every item above is how people used to enter a profession. Data entry led to operations. Tier-one support led to product specialist roles. Junior document review led to being a lawyer.
Removing the bottom rung does not just delete those jobs. It deletes the path to the ones above them, and no replacement path has appeared. A firm that stops hiring juniors this year has a senior shortage in seven years, and the decision that caused it will be long forgotten by then.
where it has not happened despite predictions
Radiology is the standard example, predicted in 2016 to be automated within five years, and radiology employment grew instead. Reading scans got faster, demand for scans rose, and accountability stayed with a licensed human.
Software engineering is the current version of the same argument. Code generation is genuinely strong, and the job was never mostly typing. What has fallen is junior hiring, which is the entry-level story again rather than replacement. Is ai replacing software developers goes into that one.
what the pattern predicts
Three traits protect work: physical presence in unpredictable space, legal accountability that must attach to a licensed person, and a relationship that is itself the deliverable. Work scoring zero on all three, screen-based, document-producing, unsigned, is where displacement lands.
The honest caveat: forecasting labour effects has a poor track record, mine included. What is observable is that whole occupations rarely disappear and specific rungs disappear quickly.
the job that is growing
Reviewing machine output, in every field it touches. Somebody reads the generated contract, checks the generated code, and decides whether the generated plan is sane. I build software that runs teams of agents, and the binding constraint is never how much they produce. It is how fast a competent person can check it.
related questions
what jobs are in danger due to ai?
screen work that produces a document and needs no signature. entry-level rungs are going first, which is a bigger problem than whole occupations disappearing.
which 5 jobs will survive ai?
no list of five is authoritative. the roles that hold up share three traits: physical presence, legal accountability, or a relationship the work cannot be separated from.
what 3 jobs will not be replaced by ai?
skilled trades, hands-on healthcare, and roles carrying legal accountability. each survives for a different reason, and the reasons are more useful than the list.
what is the 30% rule for ai?
there is no single 30% rule. the phrase covers three different claims: an automation ceiling, a productivity gain, and a hallucination budget. only one of them is useful.
spin up your first room.
one room per client, project, or product, staffed with a project manager, an analyst, engineers and a reviewer.