The model is a superhuman brain by default. Load it with a long, dumbed down spec and it hands you back exactly that: dumb. Trust it with less hand holding and you get the smarter version.

The one million token context window everyone chases is overrated. Attention cost scales with the square of your context, so it gets exponentially more expensive for barely any gain

it is very interesting to think about the first principles of delivery, and where we need to radicalize our thinking as harnesses, labs and platforms like @vorfluxai become more and more capable. it’s so easy to comply with, or stay stuck in, the conventions. instead, we need to break with those conventions. for example, you can have a 24/7 on-call engineer that acknowledges a P1/2/3 alert within seconds, delivers a solid assessment within minutes, and a few minutes later the PR is ready to be shipped. what does above really mean if you think about it for a while. and this is just one tiny aspect of delivery…

I ran AI code reviews on every PR in my repo this week.  Paid for with my Codex subscription. Most review tools bill per run, so you cap them and only point them at a few PRs…Take the token cost out and you’ll always run them. The tool that finally made that possible is Vorflux: - One model plans and build -  A model from a different lab reviews it against a fresh context - The model that writes the code is never the one that approves it - Two rival labs checking each other’s work on your task. I just recorded one PR it did to show you how fast and easy it was. It also plans, builds, and tests across your whole stack the same way, on a real cloud machine, end to end. The craziest part? The PR shows up with a recording of the thing working. I opened it, watched it run, and merged.

If your coding tool has one model supplier, you inherit their business decisions. Vorflux has never had one. @claudeai, @OpenAI , and open-weight models like kimi sit behind one dropdown or within one harness. And your plans, memory and review setup carry over when you swap.