For the last two years, whenever layoffs and AI come up in the same conversation, it almost always stays in the same place: tech. Big software companies cutting headcount, smaller engineering teams, processes that used to take ten people now handled by three.
But keeping the conversation there is a mistake of perspective. What's happening in tech is the first wave, not the only one.
What generative AI models actually automate isn't "coding". They automate tasks: writing, summarising, classifying, responding, reviewing, comparing, translating, drafting. And those tasks exist in every single industry.
Customer service, banking, insurance, admin, marketing, logistics, legal, HR, accounting: all of these sectors are built out of repetitive information-handling tasks, exactly the kind of work generative AI already performs well today. AI doesn't need to be perfect for a company to decide it needs fewer people to do the same work. It just needs to be good enough, and much cheaper.
That doesn't mean every job will disappear. It means the restructuring we've seen across tech through 2024, 2025 and 2026 is going to spread, with variations, into sectors that until now felt untouched. If you work in tech, you already know this pattern well: see what tech companies are actually paying in severance in 2026. If you work elsewhere, your turn is probably coming sooner than you think.
There's an important distinction that tends to get lost in the noise: AI doesn't replace people who know how to use it well. It mostly replaces people doing tasks AI can do as well or better, and people who haven't learned to lean on it to do more in less time.
That makes "using AI competently" the single most valuable cross-industry skill right now, regardless of your field. You don't need to become a programmer or a data engineer. You need to understand what it can do, where it fails, and how to fold it into how you work day to day: writing better, analysing faster, automating the repetitive parts, making decisions with more information in less time.
Whoever does this well isn't competing against AI. They're competing with an edge over everyone who isn't using it.
Here's an idea that probably runs against how most people were taught to think about their career: don't wait until you've been laid off to start building something of your own.
Most people only consider starting a side project, or a small business, once they're already job hunting, with time and money pressure bearing down on them. That's the worst possible moment to do it well. The best ideas and the best projects almost always start without that pressure, with room to test things, get them wrong, and learn without your next month's income depending on it.
If you're employed right now, that's exactly the moment to start. Not to quit your job tomorrow, but to build up learning, judgement, and hopefully something that works, while you still have the safety of a salary. And if you're already job hunting, don't let that stop you either: if anything, it's one of the best ways to stay active and have something to talk about in an interview besides your last job title.
The most common mistake when building something of your own is starting from the solution: "I want to build an app", "I want to launch a SaaS", "I want to create a course". That order almost never works.
The order that does work runs the other way: notice real problems, yours or people close to you, and ask why nobody has solved them well yet. It doesn't need to be big or original. Almost no successful business started from an original idea; it started from solving something that was poorly solved, slightly better than whatever existed before.
AI has radically changed the cost of getting from problem to first prototype. Today you can build a working first version of a tool, a website or an automation without knowing how to code, leaning on AI to write the code, design the interface and solve technical problems as they come up. What used to take months and a technical team can now be built by one person in days.
But it's worth being honest about the other side of this. The fact that it's so easy to build with AI without knowing how to code carries a real risk: it's just as easy to ship something badly built, without understanding what you're actually building underneath.
Poorly handled data security, business logic that breaks the moment it leaves the test case, dependencies nobody understands that break without warning. AI can move you fast, but it doesn't replace the judgement of understanding what you're building and why. Move fast, but keep asking questions, and keep reviewing what AI hands you rather than treating it as the final answer. It isn't.
Building with AI is a huge advantage if you use it with judgement, and a source of real problems if you use it without understanding what you're doing. The difference isn't in the tool. It's in who's using it.
If the uncertainty of all this change is wearing you down, you are not alone: we cover it calmly in the future work skill: learning to live with the threat of redundancy. And if you've already been through an ERE or a layoff and still don't know what you're owed, here's how severance is calculated step by step, with real examples.
If you want help preparing your next professional step (CV, interviews, search strategy, or how to position yourself in a market that's changing faster than ever), I run free 1-to-1 sessions. Message me on LinkedIn or check out the sessions.