Recently, on a construction scaffold: a big banner with the slogan "AI won't be able to paint your facade." Confident. And already wrong today.
Because that's exactly what's happening. In 2025, the first spray drones that paint building facades autonomously went into commercial use. One model coated an entire 15-story facade in under two days in the US — a job that would have taken a crew with scaffolding nearly two weeks. Around half the cost, up to 40 percent less paint thanks to millimeter-precise application, no scaffold, no fall risk. The drone that sprays paint instead of bullets — that's no longer a thing of the future.
And at a time when robotaxis already operate in everyday life in ten US cities — over 450,000 paid rides per week, fully driverless — one thing should be clear: "that'll never happen to my job" is a risky bet.
What the numbers actually say
Before the panic sets in: a sober look at what official bodies measure.
All figures in this article are as of June 2026 — AI, drone, and robotaxi numbers in particular change quickly, so treat them as a snapshot.
In its Future of Jobs Report 2025, the World Economic Forum expects around 170 million new jobs to be created worldwide by 2030 — and about 92 million to disappear. A net gain of 78 million. AI specifically accounts for roughly 11 million new and 9 million lost positions. At the same time, 41 percent of surveyed employers say they plan to cut staff where AI takes over tasks — but almost as many plan to retrain employees into other areas rather than let them go.
For Germany, the Institute for Employment Research (IAB) reaches a similar picture: total employment stays stable. By 2040, around 800,000 positions are likely to vanish — and roughly 800,000 new ones to emerge. What changes is the content of work: about 1.6 million jobs will look different than they do today. Notably, this time it isn't just simple tasks — increasingly it hits experts too, in IT, administration, even diagnostics. And already today, around 60 percent of employees in Germany use AI at work and mostly find it helpful.
In short: mass unemployment isn't in the data. A deep restructuring, however, is.
The more important question
"Will AI make me unemployed?" is the wrong question. The right one is: which side of this restructuring am I on?
Because AI almost never works entirely on its own. Even the facade drone doesn't take off by itself: someone sets it up, monitors the flight, maintains the tech, decides what gets painted when. AI shifts work around; it rarely deletes it entirely. In most places a new partnership emerges — the human leads, instructs, checks, and catches what the machine can't.
And that's exactly where your chance lies: whoever can work with AI won't be replaced — they'll be needed. Whoever understands it, directs it, and supervises it makes themselves indispensable.
How you get there
The first step is surprisingly simple — and you just took it: you're here. You're engaging with it. That sounds trivial, but it's the whole difference. If you want to know where to start, we wrote a guide to getting into AI for exactly that.
A pattern helps put it in perspective: AI becomes routine at work only once it has become everyday in private life. That's how it went with the computer. With email. With the smartphone — first a gimmick, then everywhere, then impossible to imagine life without. AI follows the same path. And the earlier you engage with it, the bigger your lead over those still waiting.
Progress can't be stopped — it arrives first where money can be made with it, and then everywhere. Probably faster than we imagine today.
No reason to panic — but every reason to start
This isn't doomsday and it isn't a sales brochure. It's the state of the facts. Yes, work is changing, noticeably, and well-trained people are affected too. That shouldn't be downplayed.
But the data also shows: at least as much new emerges as disappears. Those who move with the change don't end up worse off — they end up ahead.
So: eyes open. And then don't freeze — start. You don't have to know everything. You just have to be in.