Episode IV of AI & Docs: AI tools, automation, and an intentionally offline life

Posted on Jan 5, 2026

Fourth episode of the AI & Docs podcast series is up! In this one, Tom and I chat with CT Smith about AI workflows, building tools that don’t depend on AI, the fear of skill atrophy, living intentionally offline, and why the line between developer and writer is starting to blur.

You can watch / listen to the episode here:

Some of the things I said:


On AI dependency and losing skills

“I feel like I’m dumbing down, really. You know, it’s like—especially if I have to code something—I wonder, you know, right now I’m contributing to the toolchain at work. It’s a C# codebase. I know like the basics of C#, but I wouldn’t be able to pull the things that I pull on a monthly basis if I had access to those tools. It would be like—I don’t know—somebody removed like a lobe from my brain or something.”

“And it’s kind of scary. It’s kind of scary.”


On what AI can’t perceive

“And what’s interesting about this is that that system view—it might sound like very deterministic, like something that a machine could put together, an LLM maybe. But I don’t think they can. I don’t think they can, because the way you describe a system is always going to go through your own knowledge, your own language, your own mind, and the things that—of course, things that you know, but also things that are deeply human, like intentionality, you know, like things that people omit.”

“And also the tension—for example, I mean, all of us work every day with product management organizations that don’t really—they don’t know exactly what to do. You know, they have an idea, but they want to backtrack. And an LLM is not going to perceive those nuances because it cannot. It’s not made of flesh. It’s not organic.”


On choosing what to wrestle with

“My personal view here is that thanks to these tools, we have more freedom to choose what to wrestle with. Like, you want to read—you want to need—like, a philosophy book—okay, it’s your choice, you do it. But I don’t have necessarily—you know, I don’t have to read some slop that somebody sends me, maybe at work or outside work. Like, I will just send that to Gemini because it’s just not tempting, not interesting, you know.”


On the future of technical writing

“My take on this is that we might see some of the hyper-specialization that has been going around in tech for a while—it might disappear or slowly vanish. And I think we’ll still be technical writers, but actually, we might even drop the ’technical’ from it. We’re just going to be writers. We’re going to be folks who are whispering to machines to make them do things.”

“And the line between developer and writer is probably going to blur a bit. The only—my only concern with this is how many of us are going to be needed going forward.”


On documentation strategy

“Yeah. And, you know, documentation strategy is not only about choosing what to document, but also what not to document.”