Velocity and craft

The tools are fast. You can design a user flow in a day and build a component library in a week.

That speed makes it easy to confuse producing work with knowing whether the work is good. The person looking at the result still takes it in at the same pace. A model can build what you ask for. You still have to see what is missing, decide what matters, and know when to stop.

The trap is that some of the work these tools remove is the work that teaches you to make those decisions. How do you take the speed without skipping the learning?

What craft consists of

Production used to be a scarce skill because it was hard. As the tools make it easier, the ability to produce something tells us less about the person who made it. The difference moves to the decisions they make.

If a model gives you a competent first seventy percent, the remaining thirty percent is where your judgment matters most. You have to notice what is wrong before you can ask the tool to fix it.

Craft is what consequences leave behind. You ship a decision, it turns out slightly wrong, someone feels the problem without being able to name it, and you remember. After enough of those experiences, you start catching the wrong decision before it ships.

That is how taste forms. It is trained perception: the ability to see a difference that looks like nothing to someone else, and explain why one choice is better. You cannot build what you cannot see is missing.

Producing more work does not make every consequence arrive sooner. A pattern may reveal its problem when someone misuses it six months later. An architecture may look sound until the fourth feature has nowhere to go. Those are lessons you get from using the work over time.

The trap

Being stuck is miserable. It is also when you look at a problem closely enough to learn its details. You try a solution, find out why it fails, and change your understanding of the problem.

A tool can spare you that struggle. But if your taste is still forming, it also asks you to review its output with judgment you have not developed yet. The work that once built your judgment becomes something you can skip, while the tool still needs that judgment to be useful.

That loop will not break on its own. You have to choose some work to do slowly, even when the tool could do it for you. Spend the time where you do not yet understand the consequences of a decision. Let the speed help with the parts you already know how to judge.

Systems that multiply speed

As your taste develops, the tools become more useful. You can see what is wrong with a draft, throw out the bad parts, and keep the speed.

You can also write your judgment down. A decision becomes a component, a token, a convention, or a line on a review checklist. The next person starts with what you have already learned. This is how craft can make production faster without depending on everyone noticing the same details each time.

A system spreads mistakes just as easily. A shaky assumption can appear in forty screens, then become harder to question because people defend it as consistency. Repeating a decision does not make it right.

So give the decisions that will be reused more attention than the ones you will make once. Test a pattern before you spread it. Revisit a convention when a new use exposes its limits. A system is useful only while the judgment inside it holds up.

What to speed up

The speed is real. Turning it down proves nothing.

Use the tools where you can judge the result. Slow down where you still need to learn. Then put what you learn into the systems that make the next piece of work easier.

I want to spend less time producing work without losing the work that teaches me what is worth producing.