You take pride in the fact that, when it comes to code quality, you're simply better than AI. Give the same feature to a model and you'll spot rough edges in the draft; write it yourself and it comes out clean the first time. So today, once again, you spend two hours hand-coding a payment integration. Handing it to AI, you figure, would mean more time spent reviewing its output — and besides, it feels wasteful to outsource the one thing you do better than anyone. Two hours later, the code is clean. But the first customer interview you'd scheduled for that day has slipped to next week, again. You spent one more hour on the work you're good at, and one less hour on the work that actually decides whether your business survives.
This scene captures the trap solo founders fall into most often: holding onto tasks simply because you're good at them, and handing off only the ones you're bad at. It feels intuitive, but draw the line this way and you end up spending your most expensive hours on your cheapest work. The rule for delegation needs to be rebuilt from scratch.
"Delegate What You're Bad At" Is Only Half Right
The first answer that comes to mind when you think about delegation is: hand off what you're bad at. Founders with a marketing background outsource development and design; founders with an engineering background outsource copywriting and bookkeeping. As a way of covering your weak spots, it's not wrong. But it's only half right. This rule leads naturally to the conclusion that you should obviously keep doing whatever you're good at — and that conclusion is exactly what produced the two hours in the opening scene.
A 200-year-old trade theory fills in the missing half. It's exactly what David Ricardo laid out in 1817 as the principle of comparative advantage. The key variable isn't absolute skill — it's opportunity cost. Even someone who's good at two things can't do both at once. Spend time on one, and you give up the other. So the question isn't which task you're better at; it's what you're giving up while you hold onto it. The principle originally explained trade between nations. Swap the trading partner from a nation to an AI model, and it applies without modification.
The Better You Are at a Task, the More It Costs You to Keep It
Return to the developer whose code quality beats AI's. In absolute terms, the human wins. Hand coding to AI and quality dips slightly, and closing that gap takes review time. By the "delegate what you're bad at" rule, coding obviously stays with the founder. But comparative advantage looks at a different column too. The hour this developer spends coding is also an hour he could have spent on product strategy or a customer interview. If an hour of strategy work creates more value than the quality gap in the code costs, then handing coding to AI — and keeping only the final review — grows the business's total output more.
This is where the rule flips. The question isn't "is AI better at this than I am" — it's "what am I giving up while I hold onto it." The better you are at something, the more you want to keep doing it, but the more expensive that hour is, the more expensive whatever you're giving up in that hour also tends to be. The real price of two clean hours of coding isn't a freelancer's fee or a software subscription — it's the first customer you didn't get to meet. The tasks you cling to precisely because you're good at them are often the ones where holding on costs you the most.
Labor economics had already drawn the line comparative advantage points to — what counts as a "task you can afford to lose." David Autor and colleagues found that technology replaces routine tasks, ones whose steps can be written down as rules, while it complements non-routine cognitive tasks — judgment, creativity, reading context. What's good to hand to AI is whatever can be written down as a procedure; what should stay with you is whatever requires judgment. And routine tasks that can be written down as procedures tend to produce similar results no matter who does them — in other words, they're the tasks you can afford to lose. The line comparative advantage draws by opportunity cost and the line labor economics draws by the nature of the task land in the same place.
More evidence comes from field data. In a study by Erik Brynjolfsson and colleagues tracking 5,179 customer support agents, adopting generative AI raised productivity by an average of 14% — and the effect was largest, at 34%, among newer and lower-skilled workers. For a solo founder, the implication is clear: the payoff from delegating to AI is biggest in the areas where you're weakest. If you came up through marketing, you'll gain more by putting AI on development, design, and bookkeeping — the areas where you're effectively a novice — than on the marketing work you already know cold. What comparative advantage calls "the task you can afford to lose" and what the field data calls "the task you're bad at" point in the same direction. A 200-year-old trade theory and a recent call-center dataset, starting from entirely different places, arrive at the same conclusion — a convergence you can't see by looking at either one alone.
This division of labor doesn't so much shrink the total amount of work as change its character. As Daron Acemoglu and Pascual Restrepo have laid out, automation reduces the human share of existing tasks while simultaneously creating new ones. Hand off first drafts, and new work fills the space: setting review criteria, refining prompts, curating among multiple outputs. What shrinks is the work that can be written down as procedure; what grows is work that requires judgment. Where you hand off the task you can afford to lose, a more valuable, judgment-heavy task moves in.
Three Ways to Redraw the Line Using Comparative Advantage
First, next to each task, write down one line: what you couldn't do while you were doing this. Lay out everything you did over the past two weeks, and beside each task, note what got pushed aside. Next to "two hours of coding" you might write "customer interview"; next to "bookkeeping" you might write "next product's roadmap." Any item where what you gave up is worth more than the task itself is a delegation candidate — even if it's something you're good at. This is simply reproducing comparative advantage's arithmetic on paper.
Second, flag the tasks you're holding onto specifically because you're good at them. Tasks you've already handed off because you're bad at them aren't the problem — the problem is the ones you haven't handed off because you're good at them. On your list, mark every item tagged "I do this myself because I'm good at it," then pick out the ones whose steps can be written down as a procedure. If a task is both something you're good at and something that can be documented as a procedure, being good at it is no longer a reason to keep it. That task is your first delegation candidate.
Third, start putting AI on your weakest areas. If you're not sure where to begin delegating, look at the areas where you're effectively a beginner — as the call-center study confirmed, that's where the payoff is largest. One caveat: if you've never done a task yourself, you won't be able to tell a good output from a merely plausible one, so do it yourself at least once first, enough to build the eye that tells the two apart, before handing it off. Your first delegation in an unfamiliar area isn't about eliminating the work — it's about first becoming someone who can review it.
From Productivity to Profitability
Redrawing the line with comparative advantage isn't just about saving time. The point is to pour the time freed up by handing off low-value work into the work that builds judgment and wins customers. Two hours spent nudging code quality up by a fraction and two hours spent uncovering your first customer's real problem are not the same two hours. The first ends with a single output; the second sets the direction for every output that follows. Trading less of what you're good at for more of what's actually valuable is the first step in turning faster hands into revenue. The moment you swap "delegate what I'm bad at" for "delegate what I can afford to lose," division of labor stops being a productivity tool and becomes a profitability tool.
This series unpacks one manuscript, one installment at a time. It draws from Running a Company by Yourself, a book that pulls standard theory from accounting, economics, management, and investing — without regard for disciplinary lines — back together around the problems of solo business. The next installment covers how a simple table of delegated tasks can function as an org chart for a company with no employees, and why that dividing line has to be redrawn every quarter. If you've ever felt your business stall because you were too busy holding onto the work you're good at, start with a single sheet of paper: task on one side, what you gave up on the other.
Glossary
- The Principle of Comparative Advantage — Proposed by David Ricardo (1817). The principle that, regardless of absolute skill, both parties gain by specializing in whatever has the lower opportunity cost for them and trading. Swap the trading partner from a nation to an AI model, and it applies directly to how tasks are split between you and AI.
- The Routine vs. Non-Routine Task Distinction — Formalized in labor economics by David Autor, Frank Levy, and Richard Murnane. Technology replaces routine tasks — ones whose steps can be written down as rules — while complementing non-routine cognitive tasks that require judgment and reading context. It provides the grain along which to decide what to hand off and what to keep.
- The Task-Based Model of Automation — Proposed by Daron Acemoglu and Pascual Restrepo. It explains that automation both displaces the human share of existing tasks and creates new ones, meaning delegation changes the kind of work more than it changes the total amount.
