Professional Development – 2025 – Week 47

AI can bring productivity gains, but at the expense of learning skills and rising up the corporate ladder.
Layoffs typically affect middle-rank people; however, with AI where are you going to get the pipeline of talent to replace them?
You make senior employees by doing the job alongside someone who knows more.
The process for building skills involves doing something that’s close to the edge of your capability alongside someone who is an expert (i.e., novice + experts). Eventually you find someone learning from you as you become the expert.
AI tools allow experts to move faster, so companies aren’t incentivized (time, money) to preserve the novice-expert bond.
AI can’t currently take on tasks that require more interpersonal skills or sensitive judgment.
The key skill to learn is how to learn.

There are some bugs unit tests just won’t find.
This video shows an example (in Kotlin) of mocking an HTTP server running locally. Emily calls this a narrow integration test.

It’s inherently easier to write tests as your developing than to do it after the code is already built.
Canon TDD involves several micro-skills: design an executable test case that fails, write code to make the test pass, refactor.
TDD is hard to adopt because many of these micro-skills — including good design — take time to build up.
Clarifying the intent = explain what the behavior should be
In Hartman’s Green Path for skill acquisition, the “SOLO” step means you can work independently as long as you know most of the micro-skills to accomplish the task. If your task needs many unfamiliar micro-skills, you get stuck. Pairing helps get you unstuck.
Katas help develop micro-skills. They’re simpler problems which gives your brain capacity to figure out problems in unfamiliar ways. Over time these skills become habits.
“Test last” involves fewer micro-skills (and these are ones you already know).

Geoff Mazeroff