How we describe AI impact

What the three levels mean

Each career gets one answer to the question "How much will AI change this job?" The level is about how the work changes, not whether the job will still exist.

Where the levels come from

We use two public research studies. Both look at thousands of everyday work tasks and group them by job.

• OpenAI, "GPTs are GPTs": how many of a job's tasks AI tools could help speed up. This decides the starting level. See the study ↗
• Anthropic Economic Index: how much people already use AI for a job's tasks today. When use is very high, the level moves up one step. See the data ↗

Some careers, like Startup Founder, don't match any job in the research. For those, we use a few similar jobs instead and say so on the page.

How the text was written

The overview and the three lists on each career were drafted by an AI model, using the research above plus what we already know about the career. Every entry is checked automatically for length and alarming language, and every entry was also read by a person. If something looks wrong, it may be, and we'd like to fix it.

What predictions can't tell you

• Nobody knows exactly how fast AI will change any job. These are informed guesses, not promises.
• The research looks at whole jobs across the United States. Your town, your employer, and the kind of work you choose can be very different.
• A job that changes "a lot" is not a job to avoid. It often means new tools and new kinds of work, and people who learn those tools early do well.
• The skills that matter most, like caring about people, solving real problems, and learning quickly, are useful in every career.