I Had the Wrong Thread
My first draft joined the bonus and the power bill with the same money. Following the dollars killed the line and left the truer thread: leverage.
Process note — analytical backstage for Samsung’s $400,000 Bonus, and the $4,000 One.
My first draft had a closing line that I almost published. It joined the two halves with the same money, and it was the most shareable sentence in the draft. It was also wrong.
The piece runs two stories in one boom: a Samsung chip worker whose bonus runs toward $400,000 at the one spot in the AI economy where the people doing the work had a hand on the valve, and a household in Ohio that opens a power bill that went up because of data centers it never voted on. My first draft said those outcomes came out of the same AI pot.
Once “the same money” was on the page, I had to follow the dollars, and they don’t meet. The Samsung bonus is a slice of the division’s profit, paid out after the chips sell, which means it lands on shareholders and on the budget for the next factory. The power bill is a different pot: PJM’s capacity auction added $13.77 billion to customer bills across 13 states. I went looking for the dollar that travels from the chip worker’s contract to the ratepayer’s bill. There isn’t one. Nothing moves from one story to the other.
What the two halves actually share is the cause. One buildout hit two bottlenecks. That part is symmetric, and the piece is allowed to compress it. What is not symmetric is the direction. At the bottleneck labor could choke, labor took a cut of the upside. At the bottleneck nobody can choke, the cost just shows up. “Same money” would have flattened the one fact that makes the story worth telling: the outcomes run opposite because the leverage runs opposite.
So the thread isn’t dollars. It’s the question each bottleneck asks, which is who here can say no. The chip workers had a union and a product nobody could source elsewhere. The ratepayer has a billing address. Same boom, and the answer to that one question sorts who captures the upside from who absorbs the cost.
The most shareable cross-domain line is usually the one the chain doesn’t support. Write the mechanism and follow the dollar. If it holds, compress; if it doesn’t, name who captures and who absorbs.
Paid-tier subscribers get this in every issue: the analytical moves behind the piece, including the ones that didn’t survive contact with the primary sources. This week is a freebie for all subscribers.