Big Tech is spending $700 billion on AI infrastructure this year. They're building the agents. But they're not building what makes those agents useful. That's our job.
I run a company that helps people get rid of stuff they don't need anymore. Furniture, appliances, the random pile in your garage you've been avoiding for three years. We figure out whether it should be sold, donated, recycled, or hauled away — and then we make it happen.
Last week, I started rebuilding our AI so that other AI agents can hire it.
What Actually Happened Last Week
On March 2nd, Santander and Mastercard completed Europe's first live payment executed entirely by an AI agent. Not a demo. Not a pilot. A real transaction where an AI initiated and completed a payment on behalf of a human, in a regulated banking environment.
Meta acquired Moltbook — a social network where only AI agents can post. They're building the identity layer for an economy where AI agents hire each other.
Here's What They're Building
It's basically the internet stack, rebuilt for machines. Identity with Agent Cards and registries. Communication with MCP and A2A protocols. Discovery through agent marketplaces. Commerce via x402 and stablecoins.
OpenAI, Anthropic, Google, Microsoft, AWS, and Block are all co-governing these protocols through the Linux Foundation's Agentic AI Foundation. Six competitors. One shared infrastructure.
Why This Matters If You Build Real Things
Within 2-3 years, everyone will have a personal AI agent. When someone tells their agent "get rid of my old couch," the agent has a problem. It can search the entire internet in milliseconds. It cannot conjure a pickup truck.
It needs to subcontract. That's the opportunity. Your service needs to be hireable by AI agents, not just findable by humans.