Will AI Take Our Jobs? What the Employment Data Says

Key AI employment forecasts from major institutions

The most prominent predictions published between 2023 and 2025 generally say the following:

Goldman Sachs (2023): Around 300 million jobs worldwide could be automated by generative AI.
McKinsey (2023): Up to 30% of US working hours could be automated by 2030.
World Economic Forum "Future of Jobs Report 2025": New jobs created by 2030 are expected to outnumber those lost.
OECD (2024): Jobs that disappear entirely will be limited, but nearly all jobs will shift to a "humans working alongside AI" model.

Looking at these four forecasts together, a common theme appears: "the total number of jobs doesn't drop sharply — the content of jobs changes." That's the mainstream view.

How will the content of jobs change?

Fig. 1: All 4 institutions lean toward "jobs get reorganized" over "jobs disappear." The WEF projects a net gain of 78 million jobs.

Consider office work, for example — tasks like data entry, filing, and drafting reports. AI can handle entry and drafting at high speed. But "which information to report to the manager" or "how to apologize to a customer" still requires human judgment. Rather than whole jobs disappearing, the human share of work shifts toward "judgment, responsibility, people skills, and creativity."

Jobs most and least affected

The 8 examples below are representative ones that appear repeatedly in research and consulting reports.

Fig. 2: Jobs requiring people and physical presence (nursing, teaching, trades, sales) are 10–25% affected. Data-focused tasks are 70–90%.

"Reorganization" not "takeover"

The WEF's 2025 "Future of Jobs Report" predicts that by 2030, the number of new jobs created will exceed the number that disappear. New roles drawing attention include "AI engineer," "data scientist," "AI ethics officer," and "AI trainer." At the same time, physically demanding and people-centered roles in nursing, caregiving, construction, and electrical work are expected to see growing demand.

Just as YouTubers and social media marketers barely existed 10 years ago yet are mainstream today, new jobs "for working alongside AI" will keep appearing over the next decade.

But forecasts are just forecasts. Results will vary by country, industry, economic conditions, laws, and how fast companies adopt AI. The important thing isn't picking the single job that "won't disappear" — it's breaking down the job you're interested in task by task, and learning to distinguish which tasks AI handles well from which tasks humans remain responsible for.

For a designer, rapid generation of rough drafts is AI's strength, but deciding who this design communicates to and what it says is still a human decision. For a teacher, generating short quizzes is something AI can help with, but observing why a specific student is struggling takes human perception. This practice of breaking things down makes career thinking much more concrete.

Pitfalls to watch out for

3 things to remember when thinking about careers

Rather than searching for "jobs that will definitely survive," look for "jobs where you can grow stronger alongside AI." That's the more realistic framing.
Don't trust any single forecast. Different research institutions produce very different numbers on the same topic.
Don't confuse "tasks being automated" with "jobs being automated." In most cases, only some tasks within a job are replaced.

Why does this matter for your future?

During job applications and career decisions, companies and universities will increasingly look for evidence that you can "work effectively alongside AI." People who can clearly explain "which tasks I hand to AI" and "which judgments I make myself" carry real credibility in interviews. Having a broad understanding of how AI is reshaping jobs — while you're still in school — gives you a reliable compass when it's time to choose a path.

What you can start preparing now isn't just memorizing AI tool names. It's building your ability to read and understand text, work with numbers, explain things to others, and notice errors. The value lies not in producing things fast with AI — but in being able to evaluate what AI produces and fix it when needed.

What you can do starting today

3 steps for thinking about the future of work

Pick one career that interests you and break it down into a list of tasks (around 10).
Mark which tasks could be handled by AI and which would stay with humans — in different colors.
Think about what you can study or experience now to strengthen the "stays with humans" tasks, and write it down.

Summary

Rather than a sudden drop in the number of jobs, the mainstream view from major institutions is that the content of jobs will be reorganized. Data entry and simple admin tasks are highly automatable; jobs that require people and physical presence — nursing, education, skilled trades, sales — are more resilient. At the same time, new "jobs for working with AI" will emerge. Rather than being afraid of being "replaced," planning your future on the assumption of "working with AI" is the realistic approach.

Check When you read a jobs forecast, you need?