AI Efficiency Gains Just Got a Dev Lead Ordered to Cut 20% of His Team

Last Friday, I had dinner with a friend. The moment he saw me, he said: "My boss asked me to cut 20% of the team. I can't bring myself to do it."

My friend is a technical lead. Six months ago, he asked me to help build his company's AI Coding workflow from scratch. During that time, he often complained that his boss kept pushing him to measure how much AI actually improved efficiency, and he could never come up with a precise number, because the number was bad whether it was too high or too low.

I went to their company to give two lectures, and then provided remote Q&A, teaching them to go from local AI-assisted coding, to autonomous cloud execution, and then to building multi-agent teams. Although everyone said the lectures were good, I could sense they were resistant. It felt like it wasn't that they didn't want to use AI, but that there was simply no business left. If there's no business, what's the point of improving efficiency? Improving efficiency further would just be handing the boss a knife to stab themselves with.

And from my friend's complaints, since AI improved efficiency, the people below reported being more tired. They still carried the blame as before. Previously, one person managing one or two projects was enough. Now they generally had to continuously innovate new projects on top of the original ones. The pressure was high. "AI doesn't stop, so I can't stop either." It felt like being driven by AI. Before, they were just beasts of burden, but now it feels like being ridden by AI.

Although after the efficiency gains, the boss valued my friend more, my friend said every day at work he could feel the resentment from the people below, and it was very uncomfortable.

During the meal we discussed a lot. The conclusion we finally reached was that the company's business pie can't grow any bigger, and the outcome of AI efficiency gains is job consolidation, which means layoffs, unless the business grows.

From the boss's perspective, cutting costs and increasing efficiency means layoffs. A programmer's monthly salary is 15,000 RMB, so that's close to 200,000 RMB a year. If you don't use AI to improve efficiency, then I'll lay people off to make an example, to get everyone competing. If you do use AI to improve efficiency, then that proves the company doesn't need that many people.

After the conversation, my friend felt a bit better, but I fell silent. Ah, the wheels of history roll on! Has anyone's company encountered this situation? Let's support each other in the comments.

Follow my public account: AI in Plain Language