Luis | Indie Dev · AI x Web3

shipping one web3 x AI build a week, in public

- HolyReels: 40K followers on autopilot
- Ship or Quit (live on Base)

prev: $2M raised @ MintGate

[Luis | Indie Dev · AI x Web3 @jshanks_]

We raised almost $2M for our startup.

Then we spent two years proving we couldn't pick a direction.

It didn't die from the market. It died from us.

The most expensive lesson I've learned 🧵

[Luis | Indie Dev · AI x Web3 @jshanks_]

Do you remember when you joined X? I do! #MyXAnniversary

[Luis | Indie Dev · AI x Web3 @jshanks_]

Building with AI right now feels like early crypto. So much opportunity and optimism in the air.

I'll carry what I learned into the next build. And I'll finish that one. Bet on it.

Ship or quit. I'm done starting things I don't finish.

[Luis | Indie Dev · AI x Web3 @jshanks_]

My worst habit: I don't finish what I start. That habit helped kill a $2M startup.

So I built an app on Base that takes my money when I quit.

[Luis | Indie Dev · AI x Web3 @jshanks_]

Then I lived the exact problem it solves. I got "distracted" building an AI Catholic channel on TikTok, now 20k+ followers and 2M+ views.

Old me lets Ship or Quit die right there. The stake didn't let me. It worked.

[Luis | Indie Dev · AI x Web3 @jshanks_]

@PunkXBT_ I appreciate that. real-value incentives are one of the most powerful and unique features crypto offers imo. better than points or badges

funny timing, I'm revisiting Ship or Quit this week after diving into some AI video production work. yeah, let's connect, DMs are open

[Luis | Indie Dev · AI x Web3 @jshanks_]

Ship or Quit is live: https://t.co/Ub1tsHbCss

I just fixed the scariest problem in my app:

Users could “win” and still lose their money

So I rebuilt the flow:

Blockchain-first: stake returns first, DB updates after

Removed “Skip for now” (no more accidental loss)

[Luis | Indie Dev · AI x Web3 @jshanks_]

Two things I can build that are rarer together than apart: production AI systems + onchain products that hold money safely. Ship or quit.

[Luis | Indie Dev · AI x Web3 @jshanks_]

My worst habit as a builder: I don't finish things. So I built an app on the blockchain that takes my money when I quit.

[Luis | Indie Dev · AI x Web3 @jshanks_]

The hardest bug wasn't technical, it was philosophical. A user could "win" and still lose their money (DB updated before the funds moved). Rebuilt it blockchain-first: stake returns settle first, UI catches up. Money before pixels.

[Luis | Indie Dev · AI x Web3 @jshanks_]

You're not buying a clip. You're buying an engine.

Proof: HolyReels, 20K+ followers and ~2M views, posted daily, hands-off.

Renting a tool gets you a video. Owning the pipeline gets you a system.

[Luis | Indie Dev · AI x Web3 @jshanks_]

Anyone can type a sentence into Runway and get a 5-second clip.

That's not what I do. The gap between those two things is the whole business.

[Luis | Indie Dev · AI x Web3 @jshanks_]

I work a layer down.

Open-source models on my own GPUs, wired into a pipeline that turns one prompt into a finished video: scripted, voiced, captioned, published.

No app, no editor, no timeline.

Once it runs, it runs without me.

[Luis | Indie Dev · AI x Web3 @jshanks_]

The tech was never the hard part

Turning something people love into something that pays is. Still solving that in public.

But I know why I build it: my mom's faith deserved better than a slideshow, and I had the tools.

[Luis | Indie Dev · AI x Web3 @jshanks_]

My last startup raised ~$2M and died

The one I'm building now started on my mom's bookshelf 🧵

[Luis | Indie Dev · AI x Web3 @jshanks_]

That became HolyReels

One prompt in, a finished voiced captioned video out. I pointed the pipeline at Catholic content and let it run.

March: 529 TikTok followers
Today: 20,000+ and ~2M views

[Luis | Indie Dev · AI x Web3 @jshanks_]

I didn't get this from a book.

I got it from watching $2M of talent and capital bleed out from a lack of focus.

Finish the thing. Then pick the next thing.

Now I build AI systems that ship, in public. Ship or quit.

[Luis | Indie Dev · AI x Web3 @jshanks_]

We raised almost $2M for our startup.

Then we spent two years proving we couldn't pick a direction.

It didn't die from the market. It died from us.

The most expensive lesson I've learned 🧵

[Luis | Indie Dev · AI x Web3 @jshanks_]

So now I single-thread.

One direction at a time. Finished before the next one starts.

When a shiny new idea shows up mid-build (and it always does), I write it down and keep going.