Everyone tells you to pick a "safe" job. Almost nobody shows you the math. So here it is, straight from 2024 government data: the hands-on careers automation isn't touching — sorted by the years it takes each one to out-earn a high-school grad who skipped training and went straight to work. The winners might not be the ones you'd guess.
These pay you a real wage from the first day of your apprenticeship. No tuition, no student loans, and you're earning while everyone else is borrowing. Break-even is effectively immediate — you're never in the hole.
You pay for training here, but it's short — and the salary is high enough that you catch up to (and pass) the straight-to-work baseline quickly. "Break even" = years from your first day of school until your total earnings overtake a high-school grad who started working immediately.
More years up front, but the salary ceiling is high and demand is climbing fast. If you're willing to invest the time, the payback still lands — and these hold up against automation.
We don't sell you a career, so here's the part most "top jobs" lists leave out: some hands-on jobs are safe from AI and still a rough financial deal. The data has to show both.
These are the paths guidance often pushes as smart and stable. Most cost you four unpaid years and sit in the high-exposure zone for automation, per WEF and Stanford. Compare their break-even to the paid trades above.
This page is a starting filter, not a conclusion. Put any two careers side by side — salary, training, AI risk, job growth — or get a free report built around your own path. No login. No advice. Just the data.
Break-even is the number of years from your first day of training until your cumulative earnings pass a high-school graduate who skipped training and worked full-time from day one, at the 2024 BLS median of $49,000/yr ($946/week).
Paid apprenticeships (electrician, plumber, power-line, carpenter, firefighter, police) earn a wage during training and carry no tuition, so they break even almost immediately — we label them "paid from day one" rather than assigning a single year figure, since apprentice wages rise over the term.
The model uses national median salaries and typical training lengths. It deliberately excludes tuition — because the same career costs wildly different amounts depending on the school, and there is no single honest tuition number. Your real payback shifts with what you pay and where you work.
One honesty note on the contrast list: software development sits at Moderate AI risk in our data, not High — so we left it out of the "already being eaten" group. We only flag what the sources actually flag.
Sources: BLS Occupational Outlook Handbook, May 2024 (salary, training, outlook) · BLS Employment Projections 2024–34 · AI risk: WEF Future of Jobs 2025 and Stanford HAI AI Index 2025 (structural claims only). Figures are national medians for employed workers; they exclude overtime, shift differentials, and bonuses. Not career or financial advice.