Not literally. But close enough that it made me think.
I’ve been messing around with something called the Universal Commerce Protocol — UCP. I didn’t invent it. I just liked the concept, found the domain available, bought it, and started building something with my developer Pinto while the rest of the world figures out what it actually means.
That’s kind of how I operate.
And then I sat down with Claude and handed it my WordPress credentials, my social media login, my brand voice, my publishing queue.
I told it: be the best version of me.
That’s the same bet UCP is about. Exactly the same bet.
When you hand an AI agent your payment method, your store access, your procurement workflow — you’re not just automating a task. You’re extending trust to something that doesn’t have a face, a handshake, or a liability clause. You’re saying: I believe this system will act in my interest, even when I’m not watching.
I still checked what Claude posted before it went live. I still reviewed the drafts.
Maybe that’s the right amount of trust for right now. Verify, then extend.
The interesting question isn’t whether AI agents can be trusted. It’s whether we’re building the infrastructure to make trust possible — verifiable, auditable, revocable when needed.
That’s what UCP is supposed to answer for commerce. Pinto and I are just squatting on the concept until someone smarter shows up.
— Will
Leave a Reply