Episodes tagged: Season 2

Most teams bolt AI onto how they already work. A few are redesigning around it. We look at what changes when producing code is cheap — and why an AI-native team isn’t just a normal team with half the people cut.

The hope is that AI will finally help us clear technical debt. We argue it’s closer to the opposite — the fastest debt-generator we’ve ever built — and look at the new, harder-to-pay kind of debt nobody understood in the first place.

Developers love a stack argument. We look at how AI quietly makes much of it irrelevant — by collapsing the cost of working in something unfamiliar — and what genuinely still matters underneath the framework wars.

Most companies are doing something with AI — and measuring the wrong thing. We look at the adoption trap, why activity keeps getting mistaken for outcomes, and why AI just runs a broken process faster.

As building gets cheap, the bottleneck moves from “how” to “what”. We look at the rise of the product-minded engineer — someone who understands the customer, the business and the outcome, and happens to be able to build.

A fear runs under a lot of AI conversations: that experienced engineers are worth less now the machine can write code. We argue it’s close to the opposite — AI raises the floor on output but raises the ceiling on judgement far more.

If AI writes the boilerplate, tests and CRUD screens juniors used to learn on, how does anyone become senior? We look at the apprenticeship we’ve quietly automated away — and the slow, deferred bill that comes with it.

The whiteboard puzzle always tested the one mode the job never uses — alone, from memory, no help. We look at why AI finally makes that impossible to ignore, why hiring clings to it anyway, and what’s worth measuring instead.

We argue prompt engineering is already on its way out — and what’s replacing it is one of the oldest skills in software. When generating code gets cheap, deciding precisely what to build becomes the whole game.