Amazon, Google and Vibe Coding with Steve Yegge

Thursday, July 17, 2025

Join Steve Yegge as he delves into AI coding's complexities, Google's platform struggles, and the emergence of the 'AI Fixer' role in tech. Don't miss the insights!

Amazon, Google and Vibe Coding with Steve Yegge

Steve Yegge shares why Google struggles with platforms, how AI coding is deceptively hard, and why the "AI Fixer" role could be one spreading inside tech companies, in the future.

🎧 Stream the Latest Episode

Listen and watch now on YouTube, Spotify, and Apple Podcasts. See the episode transcript at the top of this page, and timestamps for the episode at the bottom.

🛠️ Brought to You By

  • WorkOS — The modern identity platform for B2B SaaS.
  • Statsig — The unified platform for flags, analytics, experiments, and more.
  • Sonar — Code quality and security for ALL code.

📌 In This Episode

Steve Yegge is known for his captivating rants and insightful posts, including the infamous “Google Platforms Rant.” With seven years at Amazon and thirteen at Google under his belt, he's back from a brief retirement, now working on AI developer tools at Sourcegraph. He's currently penning a new book titled Vibe Coding: Building Production-Grade Software With GenAI, Chat, Agents, and Beyond.

In today's episode, we discuss:

  • Google's recurring failures in platform development
  • The deceptively complex nature of AI coding
  • The emergent role of AI Fixers in engineering teams
  • Insights into Amazon's early practices before microservices
  • Reflections on his time at Grab and what he enjoyed about it
  • And much more!

✨ Interesting Quotes from the Episode

Google would not have hired 60% of its own people
Steve (11:13): “When I was at Google, I was on their hiring committee... One day they did an experiment with us that revealed we would have voted not to hire 60% of our own candidates. It was a sobering realization!”

Working with AI tools is hard, and great developers will thrive
Steve: “Nobody's born knowing how to do it... You cannot trust anything from AI. The best advice is to assign tiny tasks and track every line of code that will ultimately be committed.”

🤓 The Pragmatic Engineer Deepdives Relevant for This Episode

⏱️ Timestamps

  • [00:00] Intro
  • [04:55] An explanation of the interview anti-loop at Google
  • [09:50] Overview of the job-hunting hurdles at big tech
  • [15:48] Steve’s insights on Grab
  • [29:40] What went wrong with Google+ and Wave
  • [45:16] The emergence of the AI Fixer role
  • [59:24] Reasons people struggle with AI tools
  • [1:07:08] Steve’s suggestions for vibe coding
  • [1:30:08] Rapid-fire round

🔍 Find Steve Yegge Online

📚 Mentions During This Episode

Source: The Pragmatic Engineer