AI-CAREER
Navigating the AI Era as a Software Engineer
What AI actually changes for software engineering careers — layoffs, interviews, pay, and which bets are worth making — stripped of hype and doom.
Navigating the AI Era as a Software Engineer
There are two stories about AI and software engineering careers right now, and they’re not the same story.
The narrated story, the one you read on LinkedIn and in headlines, says AI is automating software engineering. Layoffs are framed as evidence. Salaries are framed as next. The implicit conclusion is that you should either pivot to AI engineering immediately or make peace with decline.
The real story is messier. Most of the 2023–2025 tech layoffs were a correction from the zero-interest-rate hiring boom, not a response to GPT-4 suddenly replacing engineers. Meta, Google, Microsoft, and Amazon all over-hired during COVID; margins were under pressure; capital got expensive. AI made a convenient public narrative because “we over-hired and are now correcting” doesn’t sound strategic. “AI is transforming our workforce” does.
Meanwhile, the thing that actually changed — and this matters — is the ceiling on individual engineer output. A competent engineer with Claude Code or Cursor can now ship what used to take a small team. That raises the bar on hiring. Junior work that was “implement this well-specified ticket” compresses. The premium on senior judgment, system design, and reliability goes up, not down.
What the career math now looks like
Three things shifted, and they compound:
- The floor rose. You can’t be a slow engineer anymore. AI tools mean your peers ship faster; matching that is now table stakes, not a differentiator.
- Interview depth increased. Companies assume you can write code with AI help. So live interviews test the parts AI is weakest at — trade-offs, system design under load, debugging unfamiliar code, and talking about impact coherently.
- The signal changed. Leetcode scores alone matter less. Shipping real systems, understanding why they work, and being able to articulate the trade-offs matters more.
A decision framework
When you’re deciding what to bet on — a new role, a new stack, a new domain — ask three questions:
- Does this build depth AI can’t easily replicate? Domain judgment, system design, regulated-data intuition, operational experience.
- Is this a narrative I can tell in an interview? If you can’t explain why you worked on it and what it moved, it’s not signal.
- Does this compound for 5+ years? Fundamentals (distributed systems, databases, networking) compound. The current hottest framework usually doesn’t.
The three pages linked below go deeper on each branch: the SWE-to-AI-engineer transition, why backend engineering isn’t dead, and a career strategy that holds up regardless of which AI narrative wins.
Frequently asked questions
- Is AI really taking software engineering jobs?
- The narrative and the data disagree. Most 2023–2025 tech layoffs mapped cleanly onto zero-interest-rate reversal, over-hiring during COVID, and an industry-wide margin reset — not AI replacing engineers. AI is raising the bar for what individual engineers can ship, which compresses some junior work, but hiring volume is still being driven by budgets, not model capability.
- Do I need to become an AI engineer to stay employable?
- No. The useful shift is becoming an engineer who uses AI fluently in your existing domain — better at prompts, evals, and code review — not abandoning backend or infra work to chase an AI title. AI-engineer roles are a real ladder, but they're not the only durable path.
- Are interviews going to disappear because AI can code?
- They're getting harder, not disappearing. Companies know candidates can lean on AI, so live interviews increasingly test reasoning, trade-offs, and system design — the parts AI is worst at articulating under pressure. Expect fewer pure LeetCode loops and more architecture and debugging conversations.