You’ve seen the headline. “Grindr CEO says AI is doing the work of 200 engineers.” If you’re an engineer, your stomach dropped. If you’re an executive, your eyes lit up. Both of you are being played.
Let’s start with what the CEO actually said. The company is measuring AI’s success by code shipped — lines of code, pull requests, commits. Sounds reasonable, right? Until you remember one of the oldest laws in tech: “The moment you measure engineering output by lines of code, you’ve already lost.” That’s Goodhart’s law in action. When a metric becomes a target, it ceases to be a good metric.
Developers know this. One HN commenter nailed it: “I’ve never worked anywhere where there isn’t at least 3 times more work to do than there is time and developers to do it.” The real constraint in every organization isn’t raw coding capacity — it’s the endless pile of contextual, judgment-heavy work that AI cannot prioritize. AI can generate code, but it can’t decide which code to write, or why.
So why would a CEO make this claim? Because the real chess move isn’t AI — it’s narrative. “Telling investors you replaced 200 engineers changes the multiple on growth, regardless of whether it’s true.” Headlines like this are designed to signal efficiency, jaw-dropping productivity, and future-proofing. They’re not about engineering reality; they’re about valuation. And the people who should be most scared? Not the engineers — but the ones who believe the story without asking what metric is being gamed.
If AI commoditizes coding, the scarce resource becomes taste, product judgment, and stakeholder trust. “AI can write code. It cannot write the story of why your product matters.” The engineers who survive won’t be the ones who can churn out more lines — they’ll be the ones who know what to build, and why.
This is a warning, not a celebration. If your company starts measuring your output by code volume, you’re already in a race to the bottom. The real work — the messy, human, contextual work — is what AI can’t replicate. And that’s exactly where you should focus.
FAQ
Q: Doesn't this prove AI is actually replacing engineers?
A: No. It proves that if you measure output by code volume, AI can generate a lot of code. But most real work is context, judgment, and prioritization—things AI can't do. The headline is marketing, not reality.
Q: How should I respond as an engineer or manager?
A: Focus on metrics that matter: business impact, product quality, user satisfaction. If your company uses code shipped as a proxy, you're incentivizing slop. Push back on the narrative.
Q: Maybe Grindr is right and AI really does replace engineers?
A: It's possible in narrow tasks, but the broader claim ignores the infinite backlog of work. The real constraint isn't coding—it's deciding what to build. AI can't do that. If anything, AI makes good product judgment more valuable, not less.