I save a pretty unreasonable number of links.

Most of them come from Twitter or HN, but sometimes an Instagram post or (on an off day) a LinkedIn post will sneak in. I’ll be scrolling around when I see a post that looks interesting, realize I don’t have time to give it the attention it deserves right now, and toss it into whatever read-it-later app I’m using. Then I keep scrolling, happy in the knowledge that I have successfully prevented this particular nugget of internet wisdom from disappearing forever.

A lot of coding wisdom gets passed from one generation of engineers to the next as a kind of meme: highly distilled, compressed lessons learned. Don’t rewrite the codebase. DRY. YAGNI. Optimize for readability. Respect complexity you don’t understand. These are useful shorthands that encode real, hard-won experience into a form that’s easy to remember and easy to pass along. Except that the ground has shifted under us, and because the knowledge is so highly compressed into these acronyms and concepts it’s difficult to tell which parts still apply. Like any good meme they’re optimized for transmission, not for accuracy, and nobody stamped an expiry date on them.

(Part 1 of ?)

I woke up, grabbed some coffee, and (as one does as a 2026 dev) took a look at what my agents had gotten up to overnight. Forty-odd tickets closed, an excessive amount of green checks, followed by a tidy summary of everything that had been accomplished. I could feel my spirits rising, it looked like a solid week’s worth of work, all done automatically for me overnight!

I’ve been trying to get off the frontier labs for a while now, for reasons I got into a bit in “good enough” models. The short version is that the cheap models have gotten good enough that paying frontier prices for mechanical coding tasks feels silly. My first thought was to try out the open weights models that have been coming out recently, but this turned out to be harder than I’d expected. Not because of the models themselves, but because finding somewhere decent to access them is surprisingly complicated. What follows is a haphazard chronicle of roughly a month of my poking around at various coding subscriptions, and how each poke has worked out so far.

The landscape of available models is changing super fast right now: as I write this Anthropic’s Fable is the undisputed best for sheer intelligence, but is also painfully expensive. OpenAI’s Sol is very close behind and much more reasonably priced, but still prohibitively expensive if you’re not on one of their subscription plans. At this point raw intelligence is no longer the only thing I care about though. Even relatively weak models (compared to the frontier) such as Zai’s GLM 5.2 are quite intelligent and more than enough for many coding tasks, at a way lower price.

An idea I’ve been trying to articulate recently is that the hard part about working with AI in 2026 isn’t actually getting them to do what we asked. Anyone who’s been using coding agents in 2026 knows this. In early 2025 your time was spent fixing up misunderstandings and mistakes the AI made. In mid 2026, that’s no longer the case. The agents are pretty good at doing what you ask them to do. The hard part today is figuring out how to accurately describe what you intended. The class of errors most worth worrying about today looks more like: “the agent understood me perfectly, followed my instructions to the letter, and in doing so showed me exactly how wrong my instructions were.”