I just finished reading Citrini Research’s 2028 Global Intelligence Crisis memo, and it’s been rattling around in my head all day. Most people are going to read it as some kind of macro-economic horror story, but if you’re actually in the trenches building software, it reads more like an inevitable production log of a system that's about to redline and explode.

The core of the problem is something we usually consider a "win": Efficiency.

In my world, we spend every waking hour trying to kill friction. We want code to write itself, we want APIs to be seamless, and we want users to be able to buy things without thinking. But Citrini’s memo makes a terrifyingly simple point: An economy actually needs friction and waste to survive. If you optimize the world to 100% efficiency, you don't get a utopia—you get a system where money stops moving entirely.

Take the SaaS industry. It’s a $13 trillion sector built on the fact that building your own software is a pain in the ass. You pay a half-million-dollar "subscription tax" to Salesforce or Zendesk because you don’t want to hire a dozen devs to build a custom version. But Citrini points out that by 2026, agentic coding tools are going to make "building" as easy as "buying." When a single mid-level engineer can clone a $500k platform in a weekend for $50 in API tokens, that entire industry disappears overnight. That’s a massive chunk of the economy just... gone.

Then there’s the death of the "Lazy Tax." Companies like Uber and DoorDash only exist because humans are too lazy to check for better prices. We pay for convenience. But AI agents aren't lazy. They don't care about a "cool" brand or a slick UI. They just optimize. If every consumer transaction is handled by an agent that pings 1,000 APIs to save two cents and find the fastest delivery, profit margins across the entire planet go to zero. You can't run a business on zero-margin efficiency.

The scariest part, though, is what he calls "Ghost GDP." We’re going to see productivity numbers that look insane on paper. We’ll be producing more "stuff" than ever because AI doesn't sleep or get sick. But that output is "ghost" value. A GPU cluster can do the work of 10,000 researchers, but that GPU cluster doesn't go out and buy a car, pay a mortgage, or go to a restaurant.

We’re essentially building a hyper-efficient engine for a car that has no one left to drive it. The money stays locked in the silicon, the velocity of money flatlines, and the real-world economy starves while the "productivity" charts go to the moon.

The Citrini memo isn't a "guess" about the future; it’s a look at the logic of the systems we’re building right now. We’re so obsessed with making everything "perfect" and "frictionless" that we’re forgetting that human messiness—salaries, middlemen, and even inefficiency—is actually the grease that keeps the gears of the world turning.

We might be the first generation of engineers to successfully optimize ourselves right out of a civilization.