I don't remember what I was looking for that night. I remember the couch. I remember the dent in the cushion that's been my dent for a long time. I remember the laptop balanced on a stack of mail, fan spinning because the battery is cooked, and the same loop I'd been running for weeks: open phone, scroll, put phone down, pick phone up, scroll. The kind of scrolling where you forget what app you're in. The kind where your thumb has more agency than your brain.
Then the OpenClaw reel hit the feed.
I'm not going to pretend I understood half of what was on the screen. Some guy with a terminal open, agents talking to agents, things spawning, tools calling tools, money moving on a wire I'd never heard of. I watched it twice. Then a third time. Then I sent it to myself in a DM so I could find it again, because even then I knew if I lost it in the algorithm I might not find my way back.
Something in my chest did this thing. Not excitement exactly. More like a door clicking open in a hallway I didn't know was in my house. That's the way out. That was the actual sentence in my head. Not 'that's cool' or 'I should learn that.' That's the way out. I'd been looking at a wall for a long time and somebody had just shown me a door, and the door was made out of code I couldn't read.
I had a laptop, an email account, and social media. That's it. I had never opened a terminal in my life. I didn't know what Python was beyond the snake. I didn't know what an API key was. I thought GitHub was where developers, like, posted screenshots of their code or something. I'm not exaggerating for the bit. That was my actual baseline.
First instinct, because I'd watched too many YouTube videos about people who 'made it' with hardware: I needed a Mac mini. I don't know why I decided that. The reel didn't say buy a Mac mini. Nobody told me to buy a Mac mini. I just got it in my head that real developers had Mac minis humming under their desk and if I was going to do this thing, the thing I now could not stop thinking about, I needed the box. So I went to buy one.
Sold out. Sold out at the Apple store. Sold out at Best Buy. Sold out at every refurbished site I could find at midnight on a Tuesday. The base model, the one I'd actually budgeted for, was just gone. I sat there with my card out feeling like an idiot. I had this little pile of money set aside, a real pile for me, and the universe had just told me no, not that.
A friend, when I complained about it the next day, said something offhand. Said: dude, you don't need a Mac mini, you need Claude. I'd heard the name. I hadn't used it. I'd been a free ChatGPT person, asking it to rewrite my texts and tell me jokes. He said get the Max plan. He said it would change my life. He said it like he was telling me where to get a good taco.
The Mac mini money became a Claude Max subscription. I want to mark that, because it feels like the first decision I made on this whole thing that turned out to matter. I didn't buy a box. I bought a brain.
Then I tried to clone OpenClaw.
I didn't know what cloning was. I didn't know what a repo was. I asked Claude to help me copy the thing in the reel, and to its credit it did not laugh at me. It walked me through what a terminal was, what git was, how to paste a URL. I got the code on my machine. I sat there looking at folders full of files I could not read in any meaningful sense. I asked Claude to just run it for me. It explained, gently, that running it was not the part I was missing.
And then it said the thing.
It said something like: you don't want to clone this. You want to build your own. You'll learn more, you'll own it, and the parts you build because you don't know any better are sometimes the parts other people can't build because they know too much.
I sat with that for a long time. I read it again. I read it a third time. The same way I'd watched the reel.
From that moment forward the project had a name in my head, even if the name kept changing for a few days before it stuck. MASTERCLAW. My own thing. Not a copy of the reel, not a fork of somebody else's hustle, my own pile of code that did the thing the reel did, in the way I could understand it, with my hands on every piece.
I'm still learning. I want to say that out loud at the start of this so I don't have to keep saying it. Senior devs are going to read these posts and clock ten amateur moves per paragraph. I know. I'm not pretending to be anything I'm not. I'm a guy on a couch in Sacramento who saw a door and walked at it.
The clock was already running when I started. I'm not going to get into all of it here. There's a date on my calendar in August that I'm building toward, and there's a reason every day between now and then matters. I'll get into the why of that later. For now just know: I am not doing this casually. I am not learning to code as a hobby. I am trying to build a thing that works, that earns, that stands up, before a date I can't move.
This post is chapter one. There are going to be more. I'm calling the series 60 Days, 27 Bots because that's the rough shape of what I'm trying to ship, and because I needed a number that scared me a little. Some of these posts will be wins. Most will be me on the couch at 2am asking the brain in my laptop why my code is yelling at me.
If you've ever seen something on a feed and felt a door open in your chest, what did you do about it?
Top comments (0)