Building is no longer thinking

I found this going through some junk last week: a plan for a side-project I was building.
pic use-cases, all features and MVP features, general architecture, user flow

Features used to take a long time to plan. How should it be architected? What about file structure? What kind of data do we need to store? Where? What are the tradeoffs in different approaches?
It used to take a while to implement it too. Typing, realizing changes are needed, refactoring, typing it again.

Answering each of those questions involved some level of research/exploration: think through approaches to achieve it, see how other people do it, Google references, chat over lunch, think of other approaches, rubber-duck it, call a friend.

The paper I’ve written is not very useful. In fact, I can’t even read it!
But the value isn’t in what’s there - it’s the distillation of thoughts while I write it.
As I churn through each step in the process, new existential questions come up.
Should I be building this feature in the first place?
Why am I building this?
Does it make sense considering the larger goals of the app?
What should the next feature be?
Should I be prioritizing this particular feature right now?
Do these engineering decisions make it easier to implement the next few features coming up?
Is this more complicated than it needs to be?
What will the user flow for the next feature look like with this?

And I mull over these questions while working on the current feature.

After AI, time taken to plan is MUCH lower. I don’t need to sift through files, search the internet or go through existing data. With the right context and judgement, I get comprehensive options for each of those questions in minutes, just waiting for my decision.
Time to implement is much lower too. Reviewing the code it spits out is more effort, but changes are quick. Typing is minimal.

The velocity and quality of implementation are now high.
Bur I feel my older workflow being thrown out of whack. I now need to intentionally schedule time to think AND resist the seductive allure of starting to build the next thing immediately.

The clarity of direction, the mental model of the product as it evolves through time, internalizing changing user expectations - all of which help make better subsequent decisions - shares no space with the building phase.

Sadly, building is no longer thinking.