Every decision you re-make is a decision you failed to delegate
You already know you can delegate to tools, environments, habits, and documents. But there is one delegation target so fundamental that it often goes unrecognized: rules.
A rule is a decision you made once, under good conditions, so you never have to make it again under worse ones. "I don't check email before 9 AM." "Any expense under $50 doesn't require approval." "If the build is red, nobody merges." Each of these is a pre-committed decision — a cognitive contract between your present self and every future situation that matches the pattern.
This is not about rigidity. This is about recognizing that the vast majority of decisions you face are not novel. They are the same decision, wearing slightly different clothes, arriving at your desk for the hundredth time. And every time you re-deliberate something you've already resolved, you spend judgment that could be applied to problems that actually require it.
Rules as decision compression
Gerd Gigerenzer, director of the Max Planck Institute for Human Development until his death in 2025, spent three decades arguing that simple rules often outperform complex analysis. His research program on "fast and frugal heuristics" demonstrated something counterintuitive: in uncertain environments, a three-question decision tree can outperform a nineteen-variable regression model. His team showed this in medical triage, financial forecasting, and consumer choice. The mechanism is not that simple rules are "good enough." It is that in noisy environments with limited data, fewer parameters mean less overfitting — the rule generalizes better precisely because it ignores most of the information.
Donald Sull and Kathleen Eisenhardt formalized this insight for organizational strategy in their 2001 Harvard Business Review article "Strategy as Simple Rules." They studied high-tech firms navigating fast-moving markets and found that the most effective companies did not rely on elaborate decision frameworks. They relied on a handful of rules — typically three to five — that everyone could remember and apply without committee deliberation. The rules were specific enough to guide action and general enough to allow judgment. "We only enter markets where we can be number one or two." "We never acquire companies with more than 75 employees." These weren't oversimplifications. They were compressed strategic intelligence.
The pattern is the same whether the domain is medicine, business, or personal life. A rule compresses a category of decisions into a single pre-made choice. Instead of analyzing each situation from first principles, you apply the rule. The cognitive savings compound. Every rule you establish is a class of decisions you no longer have to make.
The pre-commitment advantage
Rules are not just shortcuts. They are commitments made by a version of yourself who had time to think clearly, directed at a future version of yourself who may not.
The oldest illustration is the Ulysses contract. In Homer's Odyssey, Ulysses orders his crew to bind him to the mast before they sail past the Sirens. He knows that in the moment, he will want to steer toward the singing. So he makes the decision in advance — when his judgment is intact — and removes his future ability to override it. The rule ("I stay bound to the mast no matter what I say") is not a restriction on his freedom. It is a protection of his actual values against predictable interference.
Behavioral economists formalized this as the pre-commitment device. Retirement savings plans with early-withdrawal penalties, website blockers during deep work hours, and "no phones at the dinner table" agreements all share the same structure: a rule created in a calm, clear state that binds behavior in a compromised state. The rule works because it eliminates the decision point entirely. You don't decide whether to check your phone at dinner. The rule already decided. Your job is to follow it.
This is why rules are a more powerful delegation target than most people realize. When you delegate to a tool, the tool still needs you to decide to use it. When you delegate to a habit, the habit still needs the right environmental cue. But when you delegate to a rule, you eliminate the decision itself. The rule replaces the entire deliberation process with a predetermined output.
Why re-deciding costs more than you think
Roy Baumeister's influential 1998 research on ego depletion proposed that self-regulation draws on a limited resource — that making decisions literally exhausts the same mental reservoir used for self-control, focus, and willpower. While the original ego depletion model has faced serious replication challenges (a 2016 multi-lab study of 2,000+ participants found no reliable depletion effect), the behavioral phenomenon of decision fatigue remains well-documented in field settings. Judges grant parole at higher rates after meal breaks. Doctors prescribe more unnecessary antibiotics later in the day. Consumers make worse financial choices after long shopping sessions.
The practical takeaway survives even if the theoretical mechanism is contested: people who face many sequential decisions tend to either make worse choices or default to the easiest option. Rules short-circuit this degradation. Barack Obama explained his policy of wearing only gray or blue suits in a 2012 Vanity Fair interview: "I don't want to make decisions about what I'm eating or wearing. Because I have too many other decisions to make." Steve Jobs wore the same black turtleneck, jeans, and New Balance sneakers for the same reason. Mark Zuckerberg adopted a uniform gray t-shirt, telling a 2014 public Q&A audience: "I really want to clear my life to make it so that I have to make as few decisions as possible about anything except how to best serve this community."
These are not productivity hacks. They are delegation patterns. Each person identified a category of recurring, low-stakes decisions, created a rule that handles them, and redirected their finite cognitive resources toward decisions that actually matter. The rule does the work. The human does the thinking that only a human can do.
Building your own rule infrastructure
Not every decision should become a rule. The art is identifying which ones should. Here is a practical framework:
Candidates for rules: Decisions that recur frequently, produce roughly the same answer each time, and carry low-to-moderate consequences if the rule is occasionally wrong. Email triage, scheduling boundaries, spending thresholds, information diet choices, and response policies all fit.
Poor candidates for rules: Decisions that are genuinely novel, that involve high stakes with irreversible consequences, or that depend heavily on context that changes each time. Hiring decisions, strategic pivots, relationship conversations, and creative direction all require real-time judgment. Delegating these to rules doesn't free up your cognition — it surrenders it.
The anatomy of a good rule:
- Specific trigger. "If a meeting request arrives without an agenda" is actionable. "If I feel like the meeting might not be useful" is not — it requires the same deliberation the rule was supposed to eliminate.
- Clear action. "Decline it" or "Ask for an agenda before accepting." No ambiguity about what to do.
- Known boundary. Every good rule knows when it does not apply. "Unless the request comes from my direct manager or involves an active production incident." Without a boundary, a rule becomes dogma.
- Review date. Rules are hypotheses about recurring patterns. Patterns change. A quarterly review of your active rules prevents them from calcifying into constraints that no longer serve you.
Start with three rules. Write them as explicit if/then statements. Post them somewhere visible — a sticky note, a pinned document, a recurring checklist. Follow them for thirty days before evaluating. Most people find that well-chosen rules eliminate fifteen to twenty decisions per day while producing equal or better outcomes than ad hoc deliberation.
Rules as your Third Brain's operating policy
In AI-augmented workflows, rules take on a second dimension. A system prompt is a rule. A CLAUDE.md file is a set of rules. An agent's tool-use policy, its escalation criteria, its formatting constraints — all rules. When you write "never commit to main without running tests" in a project configuration, you are creating a Ulysses contract for your AI collaborator, binding its behavior to your values before the situation arises.
This parallel is not a metaphor. The same design principles that make personal rules effective — specific triggers, clear actions, known boundaries, periodic review — make AI system prompts effective. A vague system prompt ("be helpful and accurate") produces inconsistent behavior for the same reason a vague personal rule ("try to be healthier") produces inconsistent behavior: it requires re-deliberation at every decision point.
The most effective AI governance structures are layered rule systems. Organization-level policies set boundaries. Team-level conventions set standards. Individual system prompts set preferences. Each layer delegates a category of decisions so that the layers above and below don't have to re-decide them. This is exactly how well-designed personal rule systems work: some rules are non-negotiable (values), some are strong defaults (policies), and some are flexible guidelines (preferences).
When you design rules for yourself, you are practicing the same skill you need to design rules for any system — human or artificial — that acts on your behalf.
The rule audit: staying honest about what you've delegated
Rules have a failure mode that is the mirror image of their benefit. The same mechanism that saves you from re-deciding also saves you from reconsidering. A rule that was perfectly calibrated six months ago can become a trap if the underlying conditions have changed and you never noticed — because the rule prevented you from noticing.
This is why every rule system needs an audit cycle. Once per quarter, review your active rules and ask three questions:
- Is this rule still producing good outcomes? Check the results, not just whether the rule feels comfortable.
- Has the context changed since I created this rule? New information, new responsibilities, new constraints — any of these can invalidate a rule without sending a notification.
- Am I using this rule to avoid a decision I should actually be making? Sometimes a rule persists not because it is effective but because the deliberation it replaced is uncomfortable. "I never take meetings on Fridays" might be protecting your deep work time — or it might be helping you avoid a conversation you need to have.
Rules are decisions you made in advance. But you are also the person who gets to unmake them. The goal is not to accumulate rules indefinitely. It is to maintain a lean, current set of pre-committed decisions that genuinely serve the life you are building now — not the life you were building when you wrote them.
The next lesson, L-0534, addresses what happens when this balance tips: when you have delegated so much to rules, systems, and processes that you have lost contact with what is actually happening. The question shifts from "What should I delegate?" to "What have I delegated that I shouldn't have?" Every powerful delegation pattern carries the risk of over-delegation — and rules, because they are invisible once installed, carry the highest risk of all.