We held out as long as we could.
In all seriousness, we've been writing AI-assisted code at Boot .dev for quite awhile. It took time for the models and tooling to get good enough for real best practices to emerge, and (candidly) for us to get good enough at it to have something worth teaching.
Now we do.
95 interactive lessons. You'll build a real interpreter with OpenCode v2. It covers all the important features of modern agents, plus the hard-won tips we've learned through every day usage.. and a few guest-takes from @thdxr on the OpenCode team.
Enjoy.
While I spend most of every day working with LLMs and have managed to adapt to it so far, I would turn back the clock on this technology if it were within my power. I just see huge and diverse negative effects, while the positives are fewer, smaller, surprisingly incremental.