With AI, I became a hostage of my own ideas
Published on
Allow me to bring up a reflection (or “provocation”, for those who don’t hate that word 😅) that I’ve been having quite often lately. Because of the productivity gains brought by AI, I have the feeling that I’m becoming a hostage or slave to my own ideas. Let me explain.
Having cool project ideas is something that happens every now and then (even fun personal projects, not necessarily something professional). But until two or three years ago, what did we do when we had a nice idea? I’d write it down in my phone’s notepad or in a notebook and leave it for later. And of course, the most common thing is that life got in the way, other priorities showed up and most of those ideas died right there. We abandoned them because deep down we knew it would be A LOT of work to get that off the ground.
But now, with super powerful LLMs for code working together with harnesses like Codex and Claude Code, when I have a cool project idea, my brain automatically understands that it’s doable. And the worst part: that it’s fast. The execution barrier is gone.
To give you a sense of how bad it is, here’s a personal anecdote. About 3 months ago I set up an OpenClaw to test it and connected it so I could access it through my Telegram. Then one day, around 1:30 in the morning, I had already turned off the lights and was lying down to sleep. Phone charging, eyes closed. Suddenly, a project idea comes to me.
Before, I’d probably have ignored it or, at most, jotted it down in the notepad to look at later and gone to sleep. But what did I do? My brain remembered the OpenClaw. I grabbed the phone in the dark, opened Telegram and sent it a voice message: “Look, I have a project idea like this and that. Create a new repository on my GitHub right now. Then, start programming the visual part and tomorrow morning I’ll pick it up from that point”.
In other words, instead of reading a book or just going to sleep, I was actively working, being productive, simply because… I could.

Yes, I really did this at 1:34 in the morning.
On top of that, because of this absurd power, I often find myself developing five or six projects at the same time, literally with five windows of VS Code or Claude Code or Codex open simultaneously. While one AI thinks and generates code on one screen, I ALT+TAB to another, review, go back, orchestrate…
This brings me to a reflection I’ve already mentioned in some talks and to the folks who went in person to the Magos da IA meetup last week in São Paulo: AI really has brought an undeniable productivity gain to our lives, and I see myself doing in days what used to take months.
But isn’t this superpower making us work A LOT more than we used to? Without the barrier of execution difficulty, aren’t we becoming slaves to our own capacity to produce new cool ideas?
Or is it just me?