You think you're failing.
You're not.
You're succeeding perfectly — at the wrong thing.
------------------------------------------
You’re not stuck because you’re lazy.
You’re stuck because you’re chasing the wrong goal without even knowing it.
Here's a story that slapped me in the face — and might slap you too: 🧵
“Creativity is not a talent. It is a way of operating.”
Never really paid much attention to John Cleese beyond the obvious cultural references, but the distinction he made is powerful.
The interesting part isn’t that some people are “creative” and others aren’t. It’s that certain conditions allow creativity to emerge: play, curiosity, psychological safety, the freedom to throw around half-formed ideas without immediately having them judged, optimized or killed.
"Play for its own sake".
And the opposite is just as precise.
Put one person in the room who makes everyone defensive and the entire system collapses. People stop playing. They start protecting themselves.
This is probably why some conversations generate ten new ideas in twenty minutes while others make your brain feel like it has entered corporate safe mode.
Creativity isn’t magic.
Sometimes it’s just what intelligence does when nobody is strangling it.
I set up alerts for when my automation breaks.
Then I found a program on the same server that had been dead since April. Four months. No alert, ever.
Because it never broke. It just stopped running.
Every alert I had was watching for something going wrong. Nothing was watching for nothing happening.
Those are different problems. A crash announces itself. An absence is silent by definition — there's no error, no log line, no signal at all. You find out when you notice you haven't heard from it, which for me was four months.
So I stopped monitoring for failure and started monitoring for silence. Every successful run leaves a timestamp. A second job checks that timestamp once a day and shouts if it's stale.
Now the alert I get isn't "this broke." It's "this hasn't spoken in 26 hours."
The Decameron was written after the Black Death.
Ten people leave Florence.
They tell stories.
Some tragic. Some obscene. Some ridiculous.
Seven hundred years later, the machinery still works.
People lie themselves into sainthood.
Power asks trap questions.
Everyone sees the fiction and keeps playing along.
AI can summarize Boccaccio.
It cannot replace what Boccaccio trains.
Pattern recognition.
Moral suspicion.
Attention.
Italy was not producing “content.”
It was producing operating manuals for human nature.
I post almost nothing.
Not for lack of things to say. Every build session is potentially a post. The problem is it costs me 40 minutes to turn one into something I'd publish.
So I stopped treating content as a second job.
The post already exists — it's sitting in the git diff. It just needs extracting.
I wired a Claude Code command, /receipts. It reads a session's diff, finds the one thing worth sharing, drafts it in my voice. I review, I send.
Built it today and felt right that the first post be about the tool that'll draft the rest. The content was never the bottleneck.
The friction was.
Not no-code but close.
Built a BTC futures trading bot, a voice-to-X PWA named Chord, and a full X intelligence scraper with n8n + Airtable, while Claude did the heavy lifting on implementation.
I just knew what I wanted and debugged the gaps. The zero-to-shipped speed is the real unlock I guess
Never forget, anything you create in life through force and control, you'll need to sustain with the same force, the same control.
Don't chase wants only to find you've built your own prison.
Instead, create from a place of flow.
Create with ease and effortlessness. Because what's created from ease doesn't require effort to sustain.
What I’d tell anyone building complex n8n workflows with AI assistance:
Use the model that thinks well for system design.
Use the model that debugs well for implementation.
They might not be the same model.
Don’t be loyal to a tool.
Be loyal to the outcome.
If you just started building with n8n, this might interest you. 🧵
Built an n8n workflow today that scrapes X handles, normalizes the data, and stores it in Airtable.
Took two AI models to get there. Not because one is better. Because they’re good at different things.
Here’s what actually happened.
End result:
Working pipeline.
Form input.
Profile creation.
X post scraping via Apify.
Date filtering.
Normalization.
Deduplication.
Storage in Airtable with linked records.
Runs end to end. Cost per scrape under half a cent.