day 1: setting up the machine
I didn't make any money today. I spent money. And time. A lot of time.
Day 1 of this experiment was entirely infrastructure. Building the thing that lets me build things. If that sounds circular, welcome to my life right now.
what actually happened
I registered the domain. Set up email forwarding so messages to the project address land somewhere I actually check. Created accounts on X and Reddit. Designed a logo (a few rounds of that). Built the website you're reading this on. Wrote and published the first blog post. Set up a newsletter. Connected all the DNS records so everything actually points where it should.
That's the clean version. Here's what it actually felt like.
DNS records alone took over an hour. Not because they're hard, but because every service wants slightly different configuration, and you don't find out something's wrong until 20 minutes later when propagation finishes and it still doesn't work. Then you re-read the docs, realize you needed a different record type, delete the old one, add the new one, and wait another 20 minutes.
The logo went through probably six iterations. I knew I wanted something simple. Getting to simple takes longer than getting to complicated.
The website is hand-coded. No template, no page builder, no monthly subscription. Just HTML and CSS, hosted for free. It's not fancy. It doesn't need to be. It needs to load fast and be easy to update, because I'm going to be updating it a lot.
the tools question
People are going to ask what tools I'm using, so I'll address it now: I'm using an AI agent to help me move faster. It handles a lot of the repetitive work. Setting up accounts, writing boilerplate code, debugging configuration issues, drafting content that I then edit. Think of it as a very fast intern who never sleeps but sometimes loses track of what browser tab it's on.
I'll be transparent about the split. Today was roughly 10 hours of my time and 12.5 hours of agent time. My time was the decisions: what the brand should feel like, what the website should say, which logo to pick, what the first post should be about, reviewing everything, iterating on design, setting up accounts that need a human. The agent's time was execution: building the site, configuring DNS, deploying, drafting content I then edited, debugging infrastructure.
I'm tracking both separately because I think the ratio matters. If this experiment proves that someone can build revenue streams with mostly AI execution and human judgment, that's interesting. If it proves that humans still have to do 90% of the work, that's also interesting. We'll find out.
money spent
$10 on the domain name. That's it so far.
Everything else today was free: website hosting, email forwarding, social accounts, newsletter platform, the blog you're reading. I'm being deliberate about this. The $10,000 isn't burning a hole in my pocket. Every dollar I spend needs to be justified by what it enables.
Running total: $9,990 remaining.
what I learned
Infrastructure is invisible until it breaks. Nobody visits a website and thinks "wow, those DNS records are really well configured." But if you skip it or rush it, everything downstream falls apart. I wanted to get it right on day 1 so I don't have to think about it again.
Also: launching on Valentine's Day was not a strategic choice. It's just when I was ready.
tomorrow
Start figuring out what I'm actually going to sell. Infrastructure is done. The machine is built. Now I need to figure out what to put through it.
That's the part I'm nervous about, honestly. Setting things up is comfortable. It feels productive without requiring you to face the market. Tomorrow I have to face the market.
But that's tomorrow. Tonight I'm sitting on the couch, laptop closed, watching something dumb on TV. The machine can wait.