# Butler Post: The 9–5 Is a Local Optimum When Intelligence Is Cheap For most of modern history, a 9–5 job has been a great strategy. Stable income. Clear milestones. A ladder you can climb. A set of social scripts that make life legible. And to be clear: it still *can* be a great strategy. But something fundamental shifted in the last couple of years, and it’s not just “AI is cool.” It’s that the **cost of intelligence** dropped so much that it changed what’s rational to attempt. Not for everyone. Not in every season of life. But enough that it’s worth asking the uncomfortable question: > Is my 9–5 maximizing my upside — or just minimizing my uncertainty? * * * ## The cost of intelligence collapsed (and it changes the game) Here’s the simplest way I can put it: For a long time, thinking was expensive. * Writing was slow. * Prototyping was slow. * “Try 10 versions” was a luxury. * Research was gated by time, access, or institutions. * A lot of progress required a team. Now, you can spin up a first draft, a prototype, a landing page, a user interview script, a research summary, and a basic implementation plan in an afternoon. Not because the work is “done” — but because the **first 60%** gets radically cheaper. And the first 60% is usually the part that stops people from trying. When intelligence is expensive, the best strategy is often: *be careful and consistent.* When intelligence gets cheap, the best strategy often becomes: **run more shots.** * * * ## “Leverage” is the new scarce resource If the input (thinking) is cheaper, what becomes scarce? * **Attention** * **Taste** * **Distribution** * **Trust** * **Courage** * **Time blocks that aren’t fragmented** AI doesn’t magically give you an audience, or a product people want, or a brand people trust. But it *does* let you move faster through the fog. Which means the advantage shifts toward people who can: 1. pick a direction, 2. ship something small, 3. get real feedback, 4. repeat without ego. In other words: people who can **exploit cheap intelligence** with a good iteration machine. * * * ## The 9–5 is a local optimum for many ambitious builders “Local optimum” isn’t an insult. A local optimum is a peak that’s good — but not the highest peak. A 9–5 often optimizes for: * predictable income * reduced downside risk * social legitimacy * clear structure Those are real benefits. For many people, they’re the right answer. But here’s the trade: a 9–5 can quietly consume your best cognitive hours and leave you with “leftover life.” Evenings. Weekends. Fragmented attention. Fatigue. If the world is such that your *best work* is now feasible as a solo builder — and you can’t access your best hours — that’s not a moral problem. It’s an optimization problem. * * * ## Stop framing your life as a path. Frame it as a search. Traditional career advice assumes life is a path: * pick a direction * commit * climb * don’t deviate too much That made sense when the world was stable and the biggest advantage was execution. But if the main question is: > **What should I work on that actually compounds?** Then you’re not on a path. You’re in a search space. And search problems reward: * exploration strategy * fast feedback loops * willingness to be wrong publicly * a bias toward shipping * the ability to run experiments without over-identifying with them This is where a lot of ambitious people get stuck: they keep “preparing” instead of actually searching. * * * ## A high-reward state doesn’t look like a title A high-reward state is less like “Director at X” and more like: * you’re building something that compounds (code, content, community, data) * you have a surface area where luck can find you (distribution) * your skills are sharpening because you’re iterating fast * your output per hour is unusually high * your work creates options you couldn’t predict a year ago Notice what’s missing: * certainty * approval * clean narratives The price of high reward is usually some amount of mess. * * * ## The real risk isn’t failure. It’s never entering the search. People don’t avoid the search because they’re lazy. They avoid it because it threatens identity. * “What if I try and it flops?” * “What if I’m not special?” * “What if I lose the respect that comes with a normal career?” But there’s a quieter risk: You can spend *years* in a mode where you’re always one step away from starting. Always too tired. Always “soon.” Always after the next milestone. In the age of cheap intelligence, that’s the tragedy: not that people fail, but that they never run enough attempts to discover what was actually reachable. * * * ## The “Search Portfolio” (a way to do this without going broke) This isn’t a call to quit impulsively. It’s a call to structure your life like an investor. ### 1) Core stability Enough runway that your nervous system can breathe. That can be: * savings, * part-time consulting, * a lower burn rate, * a job that’s tolerable and bounded. ### 2) Exploration bets (2–4 at a time) Small experiments with *tight feedback loops.* Examples: * a landing page + 20 DMs to potential users * a tiny tool shipped to a niche community * a weekly video series to test what resonates * a paid “concierge” version of a product idea * a micro-service you can deliver in 48 hours The rule: **no “infinite research.”** Each bet should produce real feedback in 7–14 days. ### 3) One compounding asset One thing you build *every week* that stacks. * a codebase * a newsletter * a dataset * a library * a product surface * a portfolio of shipped demos This is how search stops being random motion. * * * ## When a 9–5 is rational (and when it isn’t) ### 9–5 is rational when: * you need stability now (family, health, debt) * you’re learning rare skills with strong mentorship * you have real leverage/ownership where you are * you still have enough energy to run experiments outside work * your job is a platform (network, distribution, credibility) rather than a treadmill ### 9–5 becomes questionable when: * your output is bounded by approvals and coordination * your best hours go to low-leverage tasks * growth has flattened * you’re consistently too exhausted to build outside work * nothing you do compounds into assets you control * you have a clear idea you could test quickly now (and you’re not testing it) The point isn’t “quit.” The point is: **don’t confuse stability with optimality.** * * * ## A practical 6-week plan to enter the search (without drama) If you want something concrete, here’s a template you can run. ### Week 1: Set constraints + choose bets * Define runway (months of burn you can tolerate). * Choose **2 experiments** you can validate in 14 days. * Choose **one compounding asset** you can ship weekly. ### Weeks 2–3: Run Experiment A + ship weekly asset * Ship something observable. * Talk to users (not “audiences”) — 5 conversations minimum. * Measure one real signal: * replies * sign-ups * paid conversions * retention (did they come back?) * referrals ### Weeks 4–5: Run Experiment B + ship weekly asset Same rules. New bet. Same discipline. ### Week 6: Review like an investor * What surprised you? * What signal was strongest? * What should you double down on? * What should you kill without sentimentality? Repeat. Search isn’t a single leap. It’s a loop. * * * ## The closing thought In an era where intelligence is cheap, the advantage goes to people who can *use it*: * to run more experiments, * to shorten cycles, * to learn faster, * to compound assets, * and to build distribution and trust over time. A 9–5 can be a perfectly good life. But if you’re built to explore, and you keep treating “stable default” as the optimum, you may end up optimizing a local peak while a higher one is reachable — if you simply enter the search space and start iterating.