Another good post from Roger Wong thinking through Anthropic’s findings on how offloading coding effort leads to understanding less:

So the AI group didn’t finish meaningfully faster, but they understood meaningfully less. And the biggest gap was in debugging—the ability to recognize when code is wrong and figure out why. That’s the exact skill you need most when your job is to oversee AI-generated output.

Inside it, a quote from the Anthropic post that resonated with me:

Cognitive effort—and even getting painfully stuck—is likely important for fostering mastery.

I wonder if part of the appeal of AI tools is the promise of “exercise without exercise,” like the vibrating belt machines of the 1950s.

Elsewhere, I found an essay about the craft of writing by Kristie de Garis:

Writing at speed privileges what arrives first. The obvious phrasing, the familiar structure, a thought that you heard somewhere before.

Also this:

A book is not retrieved fully formed from memory, or pulled up in a full bucket from some deep creative well in your body. 

The old saying goes “everyone dreams about having written a book, not about writing one.” Now we’re building software that allows people to “have written a book” and “have designed something.”

I am open (I think!) to the idea that the nature of the effort will change as tools change. But can’t see mastery arriving without effort. And I’m worried people will start mistaking prompting mastery for material mastery.