← The Signal

Nadella's "Future of the Firm," Translated for the Rest of Us

Microsoft's CEO just wrote down the most important AI idea most small companies haven't heard yet. Here's the version that applies to you.

This weekend Satya Nadella posted a long note on what he called the future of the firm in an AI-driven economy. Worth reading in full — but most of the reaction missed the part that matters if you don't run a trillion-dollar company. So let me translate.

His starting point: this shift isn't like the platform shifts before it. We used to bolt digital tools onto human work. Now, for the first time, there's a genuine two-way loop between people and machines — the system can actually learn from how your people work.

He splits a company's value into two kinds of capital. Human capital is the judgment, relationships, taste, and pattern recognition your people carry. Token capital is the AI capability a company builds and owns. The counterintuitive part — worth sitting with — is his claim that human capital gets more valuable as the AI grows, not less. Someone has to aim the machine at the right problems; point it at nothing and it just spins.

And here's the idea that should reframe how you treat every AI tool you touch: you can delegate the work, but not the learning. The understanding that accrues from doing something is the real prize, and it doesn't transfer when you hand the task to a model. The durable advantage was never the model you rent this quarter — those get better and cheaper for everyone at once. It's the loop where your company's specific knowledge gets encoded into a capability that improves every time you use it. That loop compounds. It's yours. And it's hard to copy.

Why this matters more for a twelve-person company than for Microsoft: most small businesses are doing the exact opposite. They pour their real expertise — support transcripts, sales calls, hard-won judgment — into tools they don't own, which absorb it and sell a sharper version back to the whole market. You can feel productive while the one thing you uniquely know quietly leaks out the side.

So, the consultative version.

For a business: stop asking "which AI tool should we buy?" and start asking "what do we know that no one else does — and are we capturing it or giving it away?" The goal isn't the smartest model. It's a simple loop where your work makes your AI better and your better AI makes your work faster, on knowledge you keep. You don't need Microsoft's infrastructure to begin. You need to stop treating your expertise as exhaust.

For an individual: same principle, smaller scale. The people who'll thrive aren't the fastest prompters. They're the ones building a personal system — notes, judgment, a point of view — that the tools amplify rather than replace. If every prompt makes the model smarter and leaves you exactly the same, you're on the wrong side of this. Keep some of the learning for yourself.

The takeaway is one question: what does your company — or your career — know that the rest of the world doesn't, and are you building a loop that compounds it, or renting tools that quietly drain it?

That's the whole game this decade. Not picking the winner. Owning the loop.

That's the signal for this Monday. See you next week. — Adam