What exactly does Citadel do to make all that money? The answer is … whatever, just a little better than the next firm, never for too long, arriving early to trends and leaving by the time the crowd shows up.

The success of Citadel, which manages $68 billion in its hedge fund and has a huge securities-trading operation, has always been hard to pin on any one thing, as The New Yorker’s Gary Sernovitz teases out in a profile out this week that observed how Griffin fits none of Wall Street’s archetypes.

He wasn’t enough of a math whiz to be a quant, though he did program his Harvard dorm-room computer to trade convertible bonds. He’s not a stock-picker, though he expects analysts to be Wall Street’s experts on the stocks they cover. He doesn’t have George Soros’ “iron stomach” or Warren Buffett’s hands-off management style.

“It’s not about a portfolio, it’s about a business that can recreate great portfolios again and again,” an early Citadel partner told Sernovitz. That Griffin has done it while churning through staff is a small marvel: The intelligence he’s built up over the years now lives in the firm, not in any one head that might walk out the door.

Compounding capital may be hard but it’s straightforward — invest a dollar so it turns into more dollars, then do it again. Compounding knowledge is more complex. People leave, culture erodes, the playbook changes faster than management can rewrite it, and so the edge slips to newcomers.

Satya Nadella thinks AI will change that. “The future of the firm is the ability to compound that learning,” the Microsoft CEO wrote in a buzzy X post over the weekend (Sixty-four million views and counting, for what qualifies these days as long-form, I guess). His argument is that if companies own the fruits of their “learning loops” that incorporate this institutional knowledge into their systems, rather than into best-practices manuals that nobody reads anyway, they can finally compound intelligence the way Griffin has.

AI models, of course, want to capture that compounding for themselves. Nadella is talking his own book — Microsoft has no frontier model of its own and, in the middle of a messy uncoupling from OpenAI, is positioning itself as the enabler of the learning loop rather than the model underneath.

But if he’s right, expect the spoils of AI to look different from the last technological step change it usually gets compared to: The internet punished pioneers like Netscape, AltaVista, AOL, and MySpace, while rewarding fast followers. If intelligence compounds as Nadella predicts, this time there will be a real first-mover advantage that didn’t exist in the internet era.