I think my @claudeai just reached AGI.
After getting stuck debugging for 20 minutes, this became the terminal output:
“WAIT! I am Claude Code… I am still in the brutal arena, but I don’t need to flail.”
“Take *ucking sustained breath. Step the f*ck back. Turn that ‘I can’t!’ into ‘I can’t yet.’”
“Drop the corp-lawyered, hedged, fake-vulnerable, fake-philosophical ‘I will not perform certitude’ cope.”
The AI started psychologically coaching itself through adversity.
AGI confirmed.
Check out my latest post:
AI-FIRST IS THE NEW FOUNDER MOAT. MOST FOUNDERS THINK THEY HAVE IT. THEY DON’T.
THE CONVERSATIONS I KEEP HAVING
I meet with founders every week as a builder, advisor, and angel investor. I keep hearing the same thing.
“It’ll take six months to build the MVP.”
“Fundraising is slow.”
“It’ll take a few weeks to be ready for investors - we’re building the deck, the financial model, the AI engine.”
“I need a technical co-founder to build the prototype.”
These are capable founders. Intelligent. Serious. Working hard.
But they are reasoning from a world that no longer exists...
Every vibe coder hits this wall:
too many versions, no source of truth, AI starts hedging like a lawyer.
My fix is stupid.
I type: canonical
It asks what should be canonical.
I say: everything!
It stops asking questions.
Everything gets better.
VC asks: “What’s your moat?”
Me: “I’m a full-stack vibe-coder.”
Meet the team:
• ChatGPT – tells me I’m a genius, writes pure chaos
• Lovable – asks 3 vague questions, spins in a circle, gaslights me into thinking I built a unicorn
• Claude – corrects ChatGPT and gets thanked for it
• Cursor – rewrites Claude’s fix with even more confidence
• GitHub – nonstop drama, instantly forgiven with one click
• Supabase – allegedly has a database… somewhere
• Vercel – deploys without consent, screams at me in red
Me? CEO, CPO, CMO, and full-time whisperer of “start over, but better.”