🧱 Building Toward Acceleration

Working in Cursor has been both helpful and revealing.

On the one hand, it’s a great tool. I’m using it daily, and it’s absolutely speeding things up compared to coding from scratch. But it also shows me, in vivid detail, how manual this whole process still is. I'm copying and pasting prompts back and forth between one chat window and another. I’m applying layered prompts by hand. I’m running and rerunning tests. I’m managing branches, commits, file changes. And it’s repetitive. Like, really repetitive.

What’s striking is how little of my time is actually spent thinking. Maybe 20–30% of it, tops. The rest is just being a sort of human middleware: waiting, noticing when something finishes, clicking the next step, waiting again. And I’m not great at that. I miss when a process is done. I forget to hit the next button. I lose the thread. I paste the wrong prompt into the chat because I confused what window I was copying from. And all that lag — human lag — adds up.

Which makes the core bet of this whole system even clearer: if I can get this loop working, everything else speeds up. Like, truly accelerates.

So yeah, this stretch is frustrating. But if this works, it’s not just a cool feature, it'll change everything yet again.