Late to Orientation, Early to Everything Else
I missed the first two days of Code Platoon. Not accidentally — I was in Las Vegas for my anniversary and I chose to be there fully. The calendar overlap was real, and I made a call.
I regret nothing.
What I did instead: I ran a solo installfest from the hotel room. Downloaded the tools, worked through the pre-work assignments, got my environment mostly functional before the cohort had its first standup. When I walked into day three — which was my personal day one — the installfest anxiety that apparently ate a lot of the first two days was already behind me. I had momentum before the room knew my name.
That's not a flex. It's a data point about how I work. I don't do well entering a room cold. I do better when I've already done the initial friction-reduction on my own time.
The bigger data point from that first week: Code Platoon has gone AI-first.
Day one of the official curriculum — my day three — wasn't Python syntax or Git workflows. It was a course on LLM fundamentals, built around Anthropic's AI Fluency materials. The framework is called the 4Ds: Delegation, Description, Discernment, Diligence. How to work with AI as a thinking partner rather than letting it do your thinking for you.
It's the right call. In 2026, teaching a coding bootcamp without addressing AI workflows is like teaching a writing class without addressing word processors. The tools are here, they're in every professional environment, and the question isn't whether to use them — it's whether you're using them with any actual skill.
So I sat through an AI fundamentals course, which was fine.
Here's the honest version of that experience: I've been working with these tools for years. Language models as writing partners, code explainers, research accelerators, revision engines. I have opinions about their failure modes. I know what they're good at and where they generate confident nonsense. I've built workflows around them that I've iterated on for long enough that they feel earned.
Sitting in an intro AI course when that's your background produces a specific feeling. It's not boredom exactly. It's more like watching someone explain a tool you've been using for years in terms designed for people who haven't touched it yet. You learn a few things — the 4D framework is a genuinely clean vocabulary for something I was doing intuitively — but mostly you're observing how wide the gap is, even in a room full of people who all opted into this.
That's the real data point. The room wasn't full of skeptics. These are veterans who chose a coding bootcamp in 2026, which means they already believe technology matters. And still, the AI literacy gap was visible. The concepts landed at different depths for different people. The idea of AI as "next-token prediction" — as sophisticated autocomplete, not as a search engine or a knowledge oracle — was new for most of the room.
I don't say that to sound superior. I say it because it means something: if the gap is this wide in a self-selected group of tech-motivated adults, it's enormous everywhere else. And the bootcamp is right to treat AI fluency as a foundational skill rather than an optional add-on.
The curriculum decision to open with LLM fundamentals before syntax drills tells you something about where software development is actually going. The engineers of the next five years aren't going to be the ones who can write the most code from scratch. They're going to be the ones who can delegate effectively to AI systems, evaluate the output, identify the failure modes, and stay the human in the loop when it matters.
I already try to be that person in my writing work. Now I'm learning to be that person with an actual engineering foundation under it.
Week one framing: I'm not here to be impressed by AI. I'm here to build the layer of understanding that lets me work with it at depth. That's the distinction.