A lunch at Miller’s All Day is 20 million tokens. To grasp the implications of AI, this analogy matters. Tokens, in the context of AI, are not raw materials or parts of a process—they are the end product. Tokens represent fully realized outputs: a book, a movie, a business. They are the units of creativity, action, and operation.
For the first time in history, data centers are no longer producing tools for creators. They are not generating files to assist an author in writing a book, assets for a director to make a movie, or templates for an entrepreneur to build a business. Instead, data centers have become factories where the output is the book, the movie, the operational business. The result is not a part of the process—it is the completed work.
Consider this: 20 million tokens could represent the entirety of a novel, crafted by AI, ready for publication. Or it could be a feature-length film, conceived, written, storyboarded, and edited in silico. Or, perhaps, it is a fully operational business, complete with AI-driven customer service, logistics, marketing, and product development—all embedded within those tokens.
But here’s the cute—and sharp—point: some will spend their limited financial resources on burgers, others will buy tokens. And there, quietly but unmistakably, lies the divide. The future will not be drawn between those who work harder or save more, but between those who understand tokens as power and those who consume without ever noticing they are spending the same currency.
The tokens are already here. AI is making them abundant, but abundance will flow to those who see their potential—not just as data, but as a new kind of wealth. Some will eat burgers; others will eat futures. That’s where the line will be drawn.