The Software Engineer Divide

Same era. Same job title. One group is being laid off; another is being hired as fast as companies can find them. The difference isn’t seniority or education — it’s how fast you’ve adapted.

Block just cut half its engineering headcount. At the same time, Indeed’s data shows software engineering job listings bouncing back sharply. The headline numbers look stable. Underneath, two completely different stories are playing out.

The engineers being cut were doing work AI now handles better: traditional debugging, document review, basic testing. The engineers being hired are doing something AI can’t yet do: translating messy real-world problems into systems that actually work. Harvard research across 200 million job postings confirms it — companies using AI are posting 7.7% fewer entry-level roles, but senior and decision-making positions are nearly unaffected.

My prediction: the engineering org of the future is diamond-shaped. A PM with product vision at the top. Five to ten AI-and-architecture engineers in the middle, designing systems and managing agent quality. And a large base of AI agents handling execution — writing code, running tests, deploying, documenting. In this structure, the scarcest person isn’t the best coder. It’s the person who speaks both fluent business and fluent AI.

That’s exactly what FDEs (Field Deployment Engineers) do — and FDE roles grew by 800%+ in 2025. Traditional industries adopting AI for the first time often need their first AI-native engineer, and that’s a better entry point than most.

If you have an engineering background and can communicate with stakeholders, your profile is getting rarer, not more common. Build a portfolio with AI tools while everyone else is waiting for experience.

#SoftwareEngineering #AIJobs #CareerPivot #DiamondOrg

💬 Read more: 2025 Year in Review (English)