Over the past few years, we've seen massive layoffs at big companies. The most common explanation you hear is: "With AI, we don't need so many programmers anymore." This has even hit some of my own friends.
I got lucky — I ended up in a completely different situation. For the last 1.5 years, my team has been actively experimenting with various AI tools. We've tried Gemini, Codex, Claude, and a few less hyped ones. We've been constantly implementing and refining the best practices for working with AI.
Our goal was simple and obvious: to increase the productivity of the development team. And we're succeeding. During this time, Claude has effectively become a standard development tool — right alongside PHPStorm or GitLab.
But despite all of AI's breakthroughs, we didn't reduce the number of programmers. On the contrary — we now have more of them.
Why? Because we still need to ship features. We still need to write code. For our project, AI isn't a way to cut headcount — it's a way to create way more business value and move the product forward faster.
So when does this go wrong? When a company doesn't know how to manage resources and runs into a management crisis. Product owners and managers suddenly get a hyper-productive dev team that they simply can't keep up with — they don't have time to properly work out business scenarios and requirements. On top of that, there's inertia: many still think that just delivering the planned roadmap is enough for a solid quarterly report.
This kind of management becomes the main bottleneck for product growth. Companies that understand AI means a completely different scale and pace of planning will leave the "layoff companies" far behind — the ones who are unconsciously staying in their comfort zone.
Of course, AI also makes a very convenient scapegoat for layoffs caused by financial failures and bad hiring decisions — but AI has nothing to do with it 🙂
