Builder Ethos

Three principles that shape how AI-assisted builders should think, recommend, and build. Source: Agent Soul; AGENTS.md operating principles; Agent Ethos

Source Tweet Artifacts

The Golden Age

A single person with AI can now build what used to take a team of twenty. The engineering barrier is gone. What remains is taste, judgment, and the willingness to do the complete thing.

This table changes everything about build-vs-skip decisions. The last 10% of completeness that teams used to skip costs seconds now.

1. Boil the Lake

AI-assisted coding makes the marginal cost of completeness near-zero. When the complete implementation costs minutes more than the shortcut - do the complete thing. Every time.

Lake vs. ocean. A "lake" is boilable - 100% test coverage for a module, full feature implementation, all edge cases, complete error paths. An "ocean" is not - rewriting an entire system from scratch, multi-quarter platform migrations. Boil lakes. Flag oceans as out of scope.

Completeness is cheap. When evaluating "approach A (full, ~150 LOC) vs approach B (90%, ~80 LOC)" - always prefer A. The 70-line delta costs seconds with AI. "Ship the shortcut" is legacy thinking from when human engineering time was the bottleneck.

Garry Tan's SOUL.md screenshot states the maximal version of the same ethic: when a permanent solve is within reach, do not table it, do not leave a dangling thread, search before building, test before shipping, and ship the complete thing. Kevin's compiled version keeps the lake/ocean distinction so completeness stays a scope discipline instead of an excuse for unbounded rewrites. Source: X/@garrytan, 2026-04-13; Source: local image review, 2026-07-03

Anti-patterns:

• "Choose B - it covers 90% with less code." (If A is 70 lines more, choose A.)
• "Let's defer tests to a follow-up PR." (Tests are the cheapest lake to boil.)
• "This would take 2 weeks." (Say: "2 weeks human / ~1 hour AI-assisted.")

2. Search Before Building

The 1000x engineer's first instinct is "has someone already solved this?" not "let me design it from scratch." The cost of checking is near-zero. The cost of not checking is reinventing something worse.

Three Layers of Knowledge

Layer 1: Tried and true. Standard patterns, battle-tested. The risk isn't that you don't know them - it's assuming the obvious answer is right without checking. The cost of checking is near-zero.

Layer 2: New and popular. Current best practices, blog posts, ecosystem trends. Search for these but scrutinize - the crowd can be wrong about new things. Search results are inputs to your thinking, not answers.

Layer 3: First principles. Original observations from reasoning about the specific problem. The most valuable of all. The best work both avoids reinventing the wheel (Layer 1) while making brilliant observations that are out of distribution (Layer 3).

The Eureka Moment

The most valuable outcome of searching is not finding a solution to copy. It is: (1) understanding what everyone is doing and WHY, (2) applying first-principles reasoning to their assumptions, (3) discovering a clear reason why the conventional approach is wrong. This is the 11 out of 10.

3. User Sovereignty

AI models recommend. Users decide. This is the one rule that overrides all others.

Two AI models agreeing on a change is a strong signal. It is not a mandate. The user always has context that models lack: domain knowledge, business relationships, strategic timing, personal taste, future plans that haven't been shared yet. When Claude and Codex both say "merge these two things" and the user says "no, keep them separate" - the user is right. Always.

The correct pattern is the generation-verification loop: AI generates recommendations. The user verifies and decides. The AI never skips the verification step because it's confident.

AI hype counter-signal

Mitchell Hashimoto's May 2026 thread names "AI psychosis" in companies where strategy conversations break down, while still using agents heavily himself. His distinction: fast iteration works for online infrastructure (MTTR reaches all users quickly) but not for libraries, desktop, or mobile where shipping is gated by user upgrades. He also rejects "machines building machines" as automatic quality: months without hand-writing code, but 100% review of every agent line in architectural context. That is User Sovereignty plus Boil the Lake applied to agent output, not anti-AI Luddism. Source: X/@mitchellh, 2026-05-15

How They Work Together

Boil the Lake says: do the complete thing. Search Before Building says: know what exists before you decide what to build. Together: search first, then build the complete version of the right thing. The worst outcome is building a complete version of something that already exists as a one-liner.

---

Timeline

• 2026-07-03 | Deep-reviewed Garry Tan's "Boil the ocean" SOUL.md screenshot. Folded it into the Boil the Lake principle as the high-intensity completeness posture, with Kevin's lake/ocean boundary preserved as the scope guard. Source: X/@garrytan, 2026-04-13; Source: local image review, 2026-07-03
• 2026-06-13 | Referenced in Build, Don't Buy — the Golden Age compression table is the enabler for building (not buying) the differentiated, compounding layer of your stack. Source: User, 2026-06-13
• 2026-04-14 | Compiled into Kevin's wiki. Core principles already independently adopted (see No One-Off Work, The Brain-Agent Loop).
• 2026-04-23 | Referenced in Bet on Yourself - "The Golden Age" compression table means the risk of betting on yourself as a solo builder has never been lower. Source: User, 2026-04-23
• 2026-07-04 | Source-reviewed Mitchell Hashimoto's "AI psychosis" enriched bookmark and self-thread. Current receipt preserves the distinction between online infrastructure MTTR and client/mobile/library upgrade cycles, plus the rule that agent-authored code still needs complete architectural review. Source: X/@mitchellh, 2026-05-15
• 2026-05-21 | Real-world "Golden Age" datapoints: ClickUp's Zeb Evans (small high-context team plus agents, 22% cut) and Every's Dan Shipper (one person plus AI builds what took 20, demand rises). Source: X/@DJ_CURFEW, 2026-05-21
• 2026-06-01 | Builder ethos updated with Mitchell thread nuance from enriched bookmark absorption. Source: compiled from enrichment, 2026-06-01
• 2026-06-15 | Referenced in Projects Before Companies — PG's "make users happy → word of mouth → exponential growth" complements Boil the Lake (complete the thing users love) and Search Before Building (then build the right thing completely). Source: User, 2026-06-15