The constraint is deciding what to build and judging it

Every model release gets the same reaction, that now we can finally build the thing. And every time, most of the teams I work with find the model wasn't what stopped them. The constraint sat somewhere else the whole time.

The model is the production engine. It writes, it reasons, it acts. It doesn't decide what's worth producing, and it doesn't settle whether what came out is any good. Those 2 calls set a product's quality, and both sit on the human side. A new model changes neither of them.

You can watch it play out. A team gets a stronger model and ships more. The extra output is more of the same, because they pointed a faster engine at the same undecided backlog. Nobody had decided which items in it deserved to exist, so the faster engine built more of everything. Upgrade the engine, keep the gaps, and you get faster mediocrity.

The other half is what the model reads. Give it a strong opinion and little to read, and it produces a confident answer built from a stale doc and a number nobody ever checked. The answer reads well, so the problem is easy to miss. If you can't click from the claim to where it came from, what you have is a plausible sentence, and owning a call you can't trace is guessing with extra steps.

This changes what to invest in. Get sharper about what's worth doing. Build the source of truth the work reads from, and build the checks that tell you whether you did it. Those compound across every model release, because each new model reads from the same source and gets judged by the same checks.

The frontier won't help a team that can't tell it what's worth building or check what it built. For that team, a stronger model means faster production of work nobody chose on purpose.

So when a new model comes out, the useful first question is "were we blocked on the model at all?" Almost always the answer is no. No upgrade ships the 2 things that were blocking you, so put the time into those, deciding what to build and judging what came out.