Jobs of the future: which disappear and which emerge with AI
The World Economic Forum report forecasts 170 million new jobs and 92 million fewer by 2030. See which disappear and which emerge with artificial intelligence.
NinaKalkazu.org
Listen to this article
Narrated by Thalita
Every week a new list of jobs of the future appears, almost always alongside the opposite list: those that artificial intelligence will erase. These lists say what, but rarely say why. And without the why, it's hard to know if your job is on one of them.
Here you see what shrinks, what emerges, the logic behind both lists and how to apply it to your routine, based on the largest survey on the subject and the most recent studies.
Quick summary: the answer in 30 seconds. By 2030, employers surveyed by the World Economic Forum forecast 170 million new jobs and 92 million fewer worldwide. Mainly repetitive office functions disappear, like cashiers and data entry clerks. Technology jobs grow fastest, but in numbers the biggest growth is in delivery workers, farmers, nurses and teachers. AI replaces tasks, not entire jobs.
Index
- Jobs of the future: what the largest study says
- Why AI replaces tasks, not entire professions
- Jobs that will disappear (or shrink significantly)
- New jobs with AI: those being created
- Labor market and AI: what's already changing inside companies
- What the numbers don't yet show
- How to know if your profession is at risk
- Frequently asked questions
Jobs of the future: what the largest study says
The largest study is the Future of Jobs Report 2025, from the World Economic Forum. It forecasts that between 2025 and 2030, 170 million jobs will emerge and 92 million will disappear. The balance is positive, at 78 million, but the shift affects 22% of today's jobs (14% created and 8% eliminated).
The report surveyed more than a thousand large employers, with over 14 million workers in 55 economies. It's a snapshot of companies' plans, not a law of physics.
| What the report forecasts by 2030 | Number |
|---|---|
| Jobs created | 170 million |
| Jobs eliminated | 92 million |
| Worker skills that change or become obsolete | 39%, on average |
| Workers who will need training | 59 out of every 100 |
The most uncomfortable figure is at the end of the calculation. Of the 59 who will need training, employers believe 11 probably won't receive it. This is the group with the most threatened employment.
Why AI replaces tasks, not entire professions
A profession is a bundle of tasks. AI doesn't fire the bundle: it takes over some tasks within it. If little remains that only a person does well, the job shrinks. If AI removes tedious work and frees time for the rest, the job may even grow.
Think of a kitchen. The food processor chops onions faster than any cook, and no restaurant closed because of it: the cook moved on to handle seasoning and timing. But the employee whose only job was chopping onions has a problem.
Stanford research shows this effect. In the August 2026 version of the study "Canaries in the Coal Mine", the authors separated two uses of AI: substitution, when it does the task instead of the person, and complement, when it helps the person do better. Where AI substitutes, employment fell. Where it complements, it remained stable or rose, especially among the most experienced.
Who loses first is the beginner. The knowledge of someone fresh out of college is "from books," which AI has also read. That of someone with ten years in the field comes from experience: dealing with difficult clients and cases that break the rules. That's much harder to copy.
Jobs that will disappear (or shrink significantly)
According to the World Economic Forum, the largest losses in number of jobs come from office functions: cashiers, ticket agents, administrative assistants and executive secretaries. Those shrinking fastest in proportion are postal clerks, bank tellers and data entry clerks.
| Profession | Why it shrinks |
|---|---|
| Cashier and ticket agent | contactless payment, self-service and mobile sales |
| Administrative assistant and executive secretary | calendar, email and reports handled by AI assistants |
| Data entry clerk | automatic reading of documents and forms |
| Bank teller | banking app and digital payments do the counter work |
| Postal clerk | tracking and mailing through the internet |
Several were already shrinking before, because of ATMs and apps. AI accelerates the trend.
Why precisely these
These tasks repeat all day, follow a clear rule and have results easy to verify: either the invoice was entered correctly, or it wasn't. When you can verify effortlessly, you can train a machine to do it, and the company can trust it.
New jobs with AI: those being created
The professions growing fastest in percentage are in technology: big data specialist (who organizes and analyzes large volumes of data), fintech engineer (digital bank or financial service), AI specialist and software developer. Functions linked to clean energy and electric cars are also at the top.
There are also positions that barely existed three years ago. One is the forward-deployed engineer: the programmer who works inside the client company to get AI functioning in its actual process. It's like the technician who, instead of just selling the machine, stays at the factory until it produces. TechCrunch reported how companies began competing for these professionals in 2026.
Another is managing agents. An AI agent is a program that executes tasks autonomously, in multiple steps. At Anthropic, creator of the Claude AI assistant, 20% of employees surveyed in an internal poll said their work became "managing Claudes," according to a company report from October 16, 2026.
The surprise: who grows most in numbers isn't from technology
In absolute number of jobs, the largest forecast growth is on the front line: farm workers, delivery workers, construction workers, salespeople and food industry workers.
Close behind come care professions: nurses, social workers, counselors and elderly caregivers, driven by population aging in wealthy countries. Teachers also grow, where the young population still increases.
These functions require physical presence, human contact or decisions in an environment that changes constantly. AI helps the nurse fill out the medical record, but doesn't bathe the patient.
New professions also die: the case of the prompt engineer
In 2023, "prompt engineer" was the trendy profession. A prompt is the instruction you write for AI, and this professional would be the specialist in finding the right words.
In 2024, a study by VMware researchers showed that prompts created by the machine itself worked better than those written by people. IEEE Spectrum magazine summarized: prompt engineering died. The skill remains useful, but became part of other jobs.
In one sentence: betting on a trendy job is risky; betting on a skill that serves multiple jobs is safer.
Labor market and AI: what's already changing inside companies
At Anthropic, which published its own numbers, fewer entry-level positions already appear, more time reviewing the machine's work and more meetings. At the same time, companies investing heavily in AI are hiring more, not less.
The company uses Claude in almost everything, so it's the most extreme case:
| What changed at Anthropic | Early 2025 | Oct. 2026 |
|---|---|---|
| Beginners among technical staff | 21% | 14% |
| Beginners among technical hires | 27% | 9% |
| Technical areas in total staff | 61% | 56% |
Time in meetings per employee rose 20% in one year. When AI produces lots of text and code, it falls to people to decide what's worthwhile. Human attention became the bottleneck.
Meanwhile, the Ramp survey, from an American corporate card company, found another side. Companies that adopted AI with high investment increased headcount by 10.2% in the following two years. Those that only tested tools stayed the same size.
And programming, the profession where AI can do the most tasks, keeps growing. A Federal Reserve study shows that programmer employment slowed significantly after ChatGPT's launch in late 2022, but continues rising.
What the numbers don't yet show
So far, the data don't show a wave of unemployment caused by AI. What appears is a narrower entry door for young people in functions where AI acts most. The question nobody can answer is speed: if the change is slow, there's time to adapt.
Studies still disagree. Stanford's calculates that employment of 22 to 25-year-olds in these functions is 19% below what it would be without AI. Meanwhile, a study from Germany's CESifo institute found no rise in unemployment among American recent graduates in 2026. We explain this dispute in AI and jobs: what the numbers say about the fear.
Both may be right. Fewer entry-level positions in one area don't mean more unemployed overall, because young people will work in something else. The cost appears later: if nobody hires the beginner, who becomes the experienced professional ten years from now?
How to know if your profession is at risk
Don't look at the job title; look at what you do during the day. List your tasks and ask four questions about each one. The more "yes" answers, the more that task can move to AI. The risk depends on how much of your day is in these tasks.
- Does it repeat almost identically every day? Entering invoices, filling spreadsheets, answering the same customer question.
- Does it follow a rule you can write in a manual? If an intern learns it in a week reading instructions, AI learns too.
- Is the result easy to verify? Number right or wrong, form complete or incomplete.
- Does it all happen on the computer? Tasks requiring being on location or calming someone in person are most protected.
| If you… | What to do |
|---|---|
| spend the day on tasks with 3 or 4 "yes" | learn to use AI in them before someone does it in your place, and pull decisions and people contact toward yourself |
| are starting your career | prefer functions where AI complements, and show you know how to verify and correct what it delivers |
| work in care, teaching or in-person service | your area tends to grow; use AI to cut bureaucracy |
| are in technology | jobs exist, but changed: understanding the client's business is worth as much as programming |
Frequently asked questions
Which professions will disappear with artificial intelligence?
No serious study talks about total elimination, but those shrinking most are cashiers, ticket agents, administrative assistants, data entry clerks, bank tellers and postal clerks. All depend on repetitive tasks easy to verify.
What are the jobs of the future?
In pace, technology jobs: AI and big data specialists, fintech engineers and developers. In number of positions, front-line and care jobs: delivery workers, farm workers, nurses, caregivers and teachers.
Will AI end programmer jobs?
The data say no, for now. Programmer employment grows more slowly since ChatGPT, but continues growing. The profile changed: fewer positions for beginners, more demand for those who understand the business and review AI code.
Do these forecasts apply to Brazil?
The World Economic Forum report includes 55 economies, and the logic of tasks applies in any country. Studies on young people and programmers use United States data. In Brazil, companies typically adopt technology more slowly, which tends to give more time to prepare.
To close
The most useful question isn't "will my profession disappear?", but "which of my tasks does AI already do?". Those who answer this honestly today have time to take their own work toward the side that grows. To follow these tools, check the artificial intelligence section.
