✳flâneur — a map of the web's best reading

Joe Magerramov's blog: The New Calculus of AI-based Coding

blog.joemag.dev · 2,252 words · saved by 1 readers

Here's where it gets interesting. A typical software team, even an experienced one, doesn't get things right all the time. Even with good testing and engineering practices, bugs occasionally make it through. We've all heard the phrase "testing in production." That reality is the main reason I've always believed that focusing on testing alone is not enough, and investing in blast radius and time to recovery is just as important. AI assisted code is no different, it may contain bugs even when thoroughly reviewed by a human, and I suspect the probabilities are not significantly different. However, when teams ship commits at 10x the rate, the overall math changes. What used to be a production impacting bug once or twice a year, can become a weekly occurrence. Even if most bugs get caught in integration or testing environments, they will still impact the shared code base, requiring investigation and slowing the rest of the team down. Once again, this is not just hyperbole—our team sees sig

Joe Magerramov's blog: The New Calculus of AI-based Coding Sunday, October 19, 2025 The New Calculus of AI-based Coding Over the past three months, a team of experienced, like-minded engineers and I have been building something really cool within Amazon Bedrock. While I'm pretty excited about what we are building, there is another unique thing about our team - most of our code is written by AI agents such as Amazon Q or Kiro. Before you roll your eyes: no, we're not vibe coding. I don't believe that's the right way to build robust software. Instead, we use an approach where a human and AI agen

Explore this link on the map →

related reading