Is your job the kind AI is coming for?
Type your job title. You get one number for how much of the work a machine could already do, which of your skills still count somewhere else, and the jobs you could move to without starting over. Free, and every number shows its source.
862 of 862 occupations
Find your job in the list
All 862 occupations, in one table. Ticket agents, mail sorters, bookkeepers, nurses, machine operators, teachers. Sort it by whichever column matters to you.
The jobs at the top are not the ones in the headlines. They are the ones where most of the day follows a set procedure, because that is the part a machine can copy.
O*NET 30.3 work-context descriptors. BLS OEWS, May 2025.
| Occupation | Exposure | Growth |
|---|---|---|
| Reservation and Transportation Ticket Agents and Travel Clerks | 73 | +2.8% |
| Air Traffic Controllers | 62 | +1.2% |
| Postal Service Mail Sorters, Processors, and Processing Machine Operators | 61 | −8.4% |
| Chemical Plant and System Operators | 60 | −6.1% |
| Customs Brokers | 60 | +3.0% |
Reservation and Transportation Ticket Agents and Travel Clerks
One number, and why it says that
The score runs 0 to 100. Ticket agents and travel clerks sit at 73, the highest in the whole list, because almost every task is a set procedure: look it up, check it, confirm it, print it.
Under the dial are the three measurements it came from. You can see for yourself whether they describe your day. If they do not, the score is wrong for you, and it says so rather than pretending otherwise.
A high score means a lot of the work could be automated today. Whether your employer will is a different question, and this does not answer it.
Jobs you could move to
Not a list of dream careers. Occupations that need the skills you already have, so the distance is months of learning rather than years and a new degree.
Anything that needs a licence you do not have gets pushed down, however well the rest of it matches. A ticket agent is not one course away from being a surgeon, and it would be dishonest to rank it as if they were.
What the ranking cannot tell you is whether anyone is waiting on the other side. NicheIQ reads that demand out of public conversations, which is the half of the decision this data does not cover.
And a second source of income
Across the most exposed jobs in the list, the same four things keep coming up as what is missing: learning quickly, writing clearly, working with numbers, and seeing how a system fits together. Programming is ninth. All four are learnable in months.
Learning is the slow half. The faster half is having something of your own running on the side before you need it, which is a different question from what to study.
Where to learn the missing half: Google Career Certificates, LinkedIn Learning, edX and PlanetNoCode. Some are affiliate links. It never changes what we list.