88% of 2026 AI funding went to US startups. if you're building outside that number, the play isn't to chase it. it's the opposite: build close to a customer who pays you, not close to a VC who might.
Two days in China and I haven't touched my wallet once. Everything runs through a phone tap. What strikes it's how invisible it feels. The best infra disappears.
paid for street food today at a stall with no register, no terminal, no cash box. a laminated QR code taped to a pole. the west keeps announcing the future of payments while here it´s been boring for years
i've shipped AI tools for companies and a couple of govs. here's what nobody
tells you:
- 90% of the work is deciding what NOT to automate
- "it works in the demo" and "it works in production" are different things
- the client doesn't want AI. they want to stop worrying.
stop trying to build a product people love.
build the one thing they hate doing, and do it for them.
the best business i saw this month charges $40/mo to rename and file your invoices.
that's it. no delight, no magic. just "i never think about this again."
i charge $300 for an hour of my time because it is legitimately not worth it otherwise
also - you might just actually fucking listen to me when you’ve got skin in the game
You could literally:
> Open a Polymarket account
> Deposit $20
> Copy one Chinese quant's wallet
> Let his BTC simulation pick trade
> Sleep through the 7,500% profit bet
> Wake up with $166,000
Start doing it right now: https://t.co/MD25OOSmzS
If your OpenClaw agents don’t go to the gym,
something is wrong with your setup.
Just added a gym to the 3D office.
When agents are learning or developing new skills,
they go train.
Even AI engineers need leg day. 🏋️