600 days: what I'm doing about The Last Economy

Emad Mostaque ran a macro hedge fund, then founded Stability AI. On August 22, 2025 he published The Last Economy, a free book that says we have "roughly one thousand days" before the rules of the economy lock in. In the same breath he writes, "It could be 800. It could be 1,200."

A thousand days from August 22, 2025 lands on May 18, 2028. Today is September 26, 2026, which makes it day 400, with 600 left. I finished the book on September 1, day 375, so more than a third of his window was gone before I started planning. Whatever I do has to fit inside two winters.

What the book says, in plain words

The core idea is that intelligence is moving from something people sell by the hour to something companies can own and copy. Mostaque calls it the Intelligence Inversion.

The strongest idea in the book is what he calls the Metabolic Rift. For 10,000 years every worker was a body that needed food, sleep and shelter. AI and robots need electricity. When machines took over muscle work, people moved to thinking work. This time the machine does the thinking, and there's no obvious third thing to move to.

He says the old system will settle into one of three futures. The default is Digital Feudalism: five platforms own the models and everyone else lives on a basic income inside a comfortable cage. The fear path is the Great Fragmentation, where every country walls off its own AI. Then there's Human Symbiosis, the one he wants, where everybody gets their own AI agent.

He thinks the good version starts in one small place that works so well other people copy it. I put his question on my Hope page:

Where do we build the first seed?

What I believe

The direction. The early data backs up the part he said would break first: the bottom rung of the job ladder.

Research from the Stanford Digital Economy Lab found that since late 2022, workers aged 22 to 25 in the jobs most exposed to AI have been losing ground, while young workers in less exposed jobs kept growing. Experienced workers show no gap like that. It's coming from companies hiring fewer young people, and the authors say they don't see economy-wide job loss yet.

One line in that paper drives my whole plan: jobs that run on written-down knowledge are shrinking, and jobs built on know-how you earn through practice and mentorship are growing. If a manual can teach it, it's getting cheaper.

What I don't buy

The date. He never shows the math behind a thousand days, which is how it can stretch from 800 to 1,200. His flagship example of a job AI makes obsolete is radiology. Geoffrey Hinton said in 2016 to stop training radiologists. Nearly ten years later, the U.S. is still short on them.

So May 18, 2028 is my deadline, and I'm planning for a slower grind of closer to ten years that hits entry-level workers first. If he's right about the speed, I want to be ready anyway.

The chapter 18 warning

Chapter 18 proposes two new kinds of money, Foundation Coins and Culture Credits. Those are the exact names of the tokens from Intelligent Internet, Mostaque's own company, which also hosts the free book. The chapter never mentions that.

It also attacks Bitcoin for capping its supply, then gives Foundation Coins the same 21 million cap and the same halving schedule Bitcoin uses. When I read it, the token hadn't launched. I'm not buying it on the strength of a book. Take the ideas and skip the token.

What I'm doing with the 600 days

Now through March 31, 2027: build a cushion and stop adding. Mostaque makes the case himself in chapter 13: without a cushion you're stuck in survival mode and can't plan past the next few months. I'm keeping my day job in cybersecurity, where I work on AI agent security, since experienced workers are holding steady in the Stanford data. Nothing new gets started unless it's on this list.

April through December 2027: turn attention into something I own. Almost everything I run now is hours for money or content for free. The biggest fix is launching the paid version of my Skool community, and the 63-lesson curriculum is already written. I'm also picking one local industry to go deep on, since that kind of know-how stays expensive. And this journal is me publishing what I think in public.

January 1 to May 18, 2028: be in position. With a cushion built and recurring income coming in, I can take equity, buy a small business, or take a founding role instead of billing hours. Mostaque argues that as everything online gets generated, the real thing gets scarce, so I'm also putting more into in-person events here in Atlanta.

Every quarter I grade myself on the four capitals from the book's Appendix C: material (money and health), intelligence (how fast I pick up skills that last), network (relationships) and diversity (income that survives a shock). Diversity is my worst grade. Content, consulting, community and speaking all stop the day I stop showing up, so it's one lane wearing four hats.

The countdown

I built a countdown to his date so I'd have to look at it. The countdown on my Builds page runs to May 18, 2028 and has a board of nine signs that would prove Mostaque right or wrong. I run my plan against it.

Mostaque ends the book on a question I keep coming back to: now that you won't have to do anything to survive, what will you choose to be? I'm still working on my answer. The plan above is how I'm spending the 600 days while I figure it out.

David West III