Andrew Norcross

Andrew Norcross didn’t get pushed out of tech because he wasn’t good enough. He got pushed out because he refused to pretend that a pattern-recognition machine is a replacement for an experienced engineer.

For 20 years, Norcross was the person companies called when the stakes were too high to get it wrong: NASA, the New York Times, GitHub, Disney. He wasn’t just writing code. He was designing the underlying architecture that made those systems work at scale. He did it without AI. He did it well. And somewhere in the last 18 months, the market stopped caring.

What makes this conversation worth sitting with is that Norcross isn’t bitter. He’s clear-eyed. He spent 10 years in finance before software, managed $2 billion in assets, and watched subprime happen from the inside before walking away from that industry after the 2008 crash. When he looks at the current AI economy, massively subsidized and propped up by VC leverage with no realistic path to profitability, he sees the same movie. “Quadruple your AI costs,” he says, “and tell me it’s saving money.”

He won’t use AI to build things, not because he’s stubborn, but because he believes you can’t ship software you don’t understand. When something breaks in production, and something always breaks in production, you have to be able to answer for it. AI-generated code can’t tell you why it did what it did. An engineer who only reviewed it can’t either.

In the meantime, he’s building fences and painting houses in Florida, doing referral-only handyman work while he waits for the correction he’s certain is coming. He goes to sleep sore. He’s not putting money in the wrong pockets. For now, that’s enough.

In this episode, you will hear:

  • How Norcross built the architecture for NASA.gov, the New York Times, GitHub, and Disney, and why that work has disappeared
  • His direct experience with bad AI-generated code 
  • Why the AI investment bubble looks a lot like subprime mortgage lending, and why the math doesn’t math
  • His reminder that AI is not intelligent, it’s a pattern-recognition machine, and why the word choice matters
  • Why no machine will ever account for the chaos and unpredictability of human behavior
  • His pivot to referral-only handyman work
  • A frank conversation about college vs. trades
  • Why he gave up on long-term career planning entirely
  • Why going to sleep sore from physical labor feels better than putting money in the wrong pockets

Resources from this episode

Follow and Review:

We’d love for you to follow us if you haven’t yet. Click that purple ‘+’ in the top right corner of your Apple Podcasts app. We’d love it even more if you could drop a review or 5-star rating over on Apple Podcasts. Simply select “Ratings and Reviews” and “Write a Review,” then add a quick line with your favorite part of the episode. It only takes a second, and it helps spread the word about the podcast.

Category: Podcast