I've been deep in vibe coding lately.
VIBE CODING: using artificial intelligence, prompted by natural language, to code. By it, I built an MVP of launchyourletter.com in a weekend.
That would have taken me two months with a developer. Maybe more.
But something I noticed in the process is that building is addictive. And without a frame, you go lopsided. Everything pointing one direction until something snaps.
So I landed on three buckets.
WHY IT MATTERS
Most creators I know are running one of two failure modes: building for everyone else until they lose themselves, or building only for themselves until they wonder why nothing satisfies.
The third bucket is the one nobody talks about. It might be the most important one.
MY THREE BUCKETS
The "for others" project is launchyourletter.com.
Concept: a tool for anyone who keeps saying they'll start a newsletter and never does. Answer some questions, get a custom blueprint, leave with a plan.
Status: 90%.
Here's what 90% actually looks like. I shipped it. Showed it to people. Then went back in and found the testimonial carousel was frozen on every device. Root cause: a CSS class mismatch—JSX had one name, the stylesheet had a different one. Fixed.
Then I found that a previous AI session had invented FAKE testimonials. Complete with fake names, fake companies, fake quotes. I deleted all of them and hunted down the real ones from a previous transcript. (Karin Grasso and Jordan Wells: never existed. Ed Oyama and Brady Ross: real people who actually said nice things.)
What's still broken: close the tab and your blueprint disappears. No login, no persistence. The skill packages are good but not done-good. Working on it.
This website is fully functional
The "for me" project is the Idea Garden.
Concept: I live in Obsidian. Ideas neural-link constantly in there. But the connections I haven't made yet are still invisible. I want something running in the background—catching reflections, conversations, random 11pm thoughts—then surfacing what's already there.
Some ideas are seeds. Some are half-grown. Some are ready to harvest and I just haven't looked yet.
Status: Phase 1 finished this morning. Twenty-seven seeds planted. Zero ripe. Honestly, I don’t even know if the “system” works yet.
Phase 1 is collection and tagging: pulling from Gmail, Obsidian, LinkedIn.
Phase 2 builds neural links between loosely related ideas.
Phase 3 is to create a visualization of the garden that might look something like this one day.
Every idea starts with a Nano Banana Pro prompt
The "just for fun" project is a pet podcast.
Concept: my two dogs in a five-minute AI-generated episode about our actual life. I gave my kids the MVP and they listened to the same five minutes for hours. Showed everyone they knew. That was enough for me to keep building.
Status: 50%—log in, profile your pets, add family details, hit a button, get an episode.
What's actually under the hood: Next.js + Supabase + Claude API + ElevenLabs. I upgraded to ElevenLabs v3 this week, which means when the script says [laughs], the dog actually laughs. PLUS the app now has memory—it pulls from past episodes and works callbacks into new ones.
My wife is already sharing it to people she knows.
The cover art for my dog’s podcast
⚠️ DANGER ZONE FOR CHRISTIANS
Bucket three is the first thing people cut when life gets full. But the fun project isn't a reward for finishing the other two. Half my good ideas on launchyourletter.com came while I was playing around with the pet podcast. Kill the fun and the whole operation gets duller.
Solomon wrote about a time to plant and a time to harvest. He wasn't scheduling separate seasons but describing a full life that holds both at once. We have been wired to build in a way that serves others and brings joy
IF YOU WANT TO TRY THIS
🪣 Name your projects and sort them into buckets. Which one is for others? For you? Just for fun? If a bucket is empty, you're lopsided.
🐶 Give the fun project permission to stay fun. The moment I monetize the pet podcast, it stops being the “fun project”. Keep one project completely outside the business logic. (NOTE: this doesn't mean I won't monetize my fun projects, but I know it is time to replace it with another fun project once I cross that “business” threshold.)
📣 Give the "for others" project external accountability. Tell someone about it. Give it a deadline that matters to someone besides you. The service project needs gravity it doesn't generate on its own.
Three projects. Three purposes.
Vibe coding made all three possible at the same time. I honestly have no idea what half the coding phrases and terminology mean. I'm able to do just do things because of where AI is today
Ship what serves. Build what fills. Keep something in your back pocket for fun.
♻️ If this gave you something, forward it to one someone who's grinding without a game plan. Give them permission to keep one project just for fun.
Create dangerously,
—Payton
P.S. If you keep saying "I want to start a newsletter someday," someday just got easier. → launchyourletter.com
