The Blank Canvas Is Still the Hard Part

AI tools have made building cheap. I ran 11.5 billion tokens through Claude Code in under eight weeks, and I have skills set up for most of the heavy lifting. None of that helps when I sit down with nothing to build. The blank canvas is still the hard part.

What changed it for me was to stop hunting for ideas. Every time I went looking for a good app idea, I ended up with something generic that I lost interest in within a week. The ideas that stuck came from three places:

• Something that already bugs me.
• Something that makes me laugh.
• Something I know better than most people.

Let's Touch Base hits all three. I've sat through a lot of icebreakers, and most of them go the same way: a login nobody remembers, a link IT blocks, and twenty minutes gone before anyone plays. I'm also competitive, and I think office jargon is funny. So the app is 19 quick games named after the phrases we all love to hate, like Quick Sync, Put a Pin in It and AI Can't Take My Job, and every game ends with a mock performance review. My day job shaped it too. One iPhone hosts. Players join from their phone's browser with a QR code or, on Wi-Fi, a four-digit meeting ID. Nobody creates an account, and on Wi-Fi the game stays on the local network. That's the version I'd want to approve if I were the one in IT getting the request.

That's when building stopped feeling like work. When the jokes, the game rules and the network design all come from how I actually think, the product doesn't feel forced. Problem solving and having fun turn into the same activity. If you're stuck staring at an empty repo, skip the brainstorm and start with whatever annoys you.

Keep thinking. Keep building.