Is AI Helping Engineers — Or Making Things Worse?

Season 1, Episode 10 • 25:49

Is AI Helping Engineers — Or Making Things Worse?

Welcome back to I'm a Software Engineer ~ What Next? The podcast for devs figuring out their next move. Hosted by James Wilson and Matt Sinclair, we talk honestly about what it takes to build a meaningful career in tech.

In this week's episode, James and Matt take a look at AI-based code generation tools. Products like GitHub Copilot, Claude Code, and Tidewave are everywhere now — but are they making software engineers more productive, or just causing a whole new class of problems? In this episode, we unpack how AI is changing the way engineers work, code, and think.

In this episode, we cover:

  • Why AI is both a superpower and a trap
  • The shift from imperative to declarative computing
  • How tools like Claude can help you rubber-duck a spec
  • Why engineers who refuse to adapt may fall behind
  • What hiring managers are really looking for in the age of AI

If you're a developer wondering whether AI is your best friend or worst enemy, this one's for you.

🎧 26 minutes. Zero fluff. Tune in now.

⏱️ Chapters

  • 0:00 Intro: the Dragon Ball Z Goku t-shirt
  • 0:55 The reel that blew up: is AI making the engineer's job better or harder?
  • 1:48 Different, not better: change is inevitable, and here are the two extremes
  • 3:24 The mech suit: Ripley in Aliens, 10x if trained, a world of hurt if not
  • 4:51 The candidate who won't use Copilot, and the client's nuanced answer
  • 6:36 Seduced by magic code: Matt threw away 60% of what Claude wrote
  • 8:12 Will the code get better? Problem-solving minds, and safe vs unsafe use
  • 9:56 Imperative to declarative computing: you keep the why and what
  • 11:43 Rubber-ducking a spec with Claude, and code as the design canvas
  • 13:29 What hiring managers look for: the what stage, and mission alignment
  • 15:52 Matt's 300-point Hacker News comment: why so much negativity?
  • 17:56 James builds a task list in JavaScript with ChatGPT
  • 19:10 Make it work, make it right, make it fast: know the risks
  • 20:19 Not a scam, but there are scammers: deepfakes are the tech working
  • 21:32 Yes to both: fewer engineers per problem, but the level of commodity rises
  • 23:30 Embrace it: the models plateau, but the tooling goes gangbusters
  • 24:47 Wrap-up: change brings opportunity, and send us your topic ideas

🗒️ Show Notes

🎙 Hosts

🌐 More from us

🎬