Quality is the absence of problems
Quality is a spectrum up to “perfection”
Quality relies on the organisation
Quality gets harder with scale
Quality is not necessary, but ideal
In a typical agentic workflow, high agency is the art of knowing when to delegate, when to inspect, when to stop, and when to own the result of a process.
The inner loop is where the work is done. Loops are designed to be as independent as possible. Put all quality assurances and verification inside the loop. Once you’ve designed and validated the loop itself, the only thing you have left to do is to grant autonomy by putting in place a back-pressure mechanism that acts to control the rate at which the loop is run and its scope of operation. And put humans in their rightful place, on the right decisions. Don’t treat understanding as a hand-off or a release gate, but rather as a point of decision where humans are primed to provide their insight. And then for every artifact that exists and is fed back into production and into new teams and engineers, leave behind better artifacts.
Interesting statement. Difficult to validate, as it’s mostly long term observation, and a lot can happen.
After a few cycles the company has a population of well-meaning people who do not suspect anything is off. They have only ever known the cave, and life inside the cave is good.
Generally speaking it’s best to get something 50%+ built and launch. People get caught up over perfection but never launch. Speed wins.
Fire yourself as fast as possible.
Anything meaningful takes 10+ years to build. People get bored in years 2-9, and lose focus.
Consistency in everything. If you lose it, fight like hell to get it back. People sniff it on you and want to join the winning side.
People underestimate what they can do in 10 years, but overestimate what they can do in 1 year.
Generally speaking, you are better off improving something than doing something new. Even 1% better can net out billion dollar outcomes in large markets.
Follow whales → If someone launched in your category and there is massive demand exploding, you can just swim behind the wake of the whale and have a 1000x easier
time winning.
A rising tide lifts all boats. Markets are big enough for everyone, and your competitors advertising your category just helps you.
1 A player is better than 3 B players, and 10 C players. Hire A players, give them ownership and give them opportunity. Your team will be better than you.
In general leaders exist to uplift others. It’s okay to take credit and it’s okay to seek glory, but it takes a team to do anything and your team should be smarter than you.
Julius Caesar was the first propagandist. When he was off in Gaul conquering provinces, he would journal and send back snippets to Rome.
Referrals are the most powerful marketing channel in existence. Everything is mediocre, so when someone is singing your praises it means you outperformed. Lean in.
TL;DR: A good tool is and ought to be invisible—striving to make such tools is the goal of a toolmaker.
Agreed. But still, if one can have fun with their tool, I consider a win too, because one is spending lots of time with their favorite tool, so might as well have fun. Not everything must find best of productivity / performance.
Learning anything is a long term project, and long term projects are necessary for building a sense of control over your circumstances. Almost nothing can be deliberately and meaningfully changed within the scope of a day, but in months, certainly years, a lot of things can be made to happen.
Instead, we’re switching to this wrong-headed mindset that coding agents are like compilers. That mindset gives us permission to ship terribly written code. Agents aren’t compilers - they’re more like freshly onboarded interns. They read partial possibly slopified code, take an imprecise description of the change, and must generate a change.
Humans can’t surrender their thinking and taste to armies of interns. And being hands-on, rather than consumers, helps.
The inner loop is where the work is done. Loops are designed to be as independent as possible. Put all quality assurances and verification inside the loop. Once you’ve designed and validated the loop itself, the only thing you have left to do is to grant autonomy by putting in place a back-pressure mechanism that acts to control the rate at which the loop is run and its scope of operation. And put humans in their rightful place, on the right decisions. Don’t treat understanding as a hand-off or a release gate, but rather as a point of decision where humans are primed to provide their insight. And then for every artifact that exists and is fed back into production and into new teams and engineers, leave behind better artifacts.