Few One-Way Doors
Most decisions can be walked back. Treat reversible choices as experiments with rollback paths, not courtroom verdicts on your whole future.
Thesis
There are fewer one-way doors than fear pretends. Many decisions feel irreversible because they are emotionally loaded, public, expensive, or identity-adjacent. But most can be changed, narrowed, reversed, migrated, or corrected once reality gives feedback. Source: User, 2026-06-29
The decision-making upgrade is to classify the door before paying the anxiety tax. If it is a real one-way door, slow down and protect the downside. If it is a two-way door, move, learn, and keep the rollback path visible.
The Classification
| Decision type | Examples | Default posture |
|---|---|---|
| One-way door | Ethical line, legal commitment, trust breach, irreversible data loss, reputation-damaging lie, safety-critical launch | Slow down. Seek review. Preserve integrity. |
| Expensive two-way door | Major architecture bet, vendor migration, pricing model, hiring plan, public launch angle | Decide with a rollback or migration path. |
| Cheap two-way door | Copy, UI direction, internal workflow, package choice, experiment, outreach angle | Move quickly. Let feedback correct you. |
| Fake door | Fear of judgment, imagined future regret, perfectionism, comparison | Do not let it govern the decision. |
The important category is the middle: expensive but reversible. These are the decisions that steal the most time because they are not trivial, but they are also not permanent. They deserve thought, not paralysis.
Why It Matters
The cost of over-treating two-way doors as one-way doors is compounding delay. You wait for certainty, but certainty often only appears after you try the thing. This is why Personal Principles (System Instructions for Myself) says to make decisions when things are only 50% clear. The missing half comes from contact with reality.
In startups, the penalty is brutal. Markets, customer language, model capability, and distribution channels move while you are still asking whether the door is locked. Speed as a Strategy only works if reversible choices are allowed to be fast.
The Rollback Discipline
Reversibility is not an excuse to be careless. It is a design requirement.
- Keep the previous state recoverable.
- Make experiments small enough to learn from.
- Write down the assumption you are testing.
- Define the point where you walk back.
- Avoid moral, legal, trust, or safety shortcuts that cannot be undone.
This is the calm version of speed. You can move quickly because you preserved the exit.
Failure Modes
- Paralysis cosplay as prudence. Calling every decision "strategic" because deciding is uncomfortable.
- Reversibility without recordkeeping. If nobody knows what changed or why, walking back becomes expensive.
- Treating trust as reversible. Some things can be apologized for but not restored. Lies, betrayal, and safety violations are closer to one-way doors.
- Sunk-cost drift. A reversible decision becomes irreversible only because ego refuses to walk it back.
- Decision-by-fear. Choosing the safer-feeling option only because the imagined downside is vivid.
Agent Implication
When an agent is helping Kevin decide, it should name:
- what kind of door this is
- what would make it irreversible
- what rollback path exists
- what feedback would justify reversing it
- what small version can be tried first
Do not inflate a reversible choice into a fake existential crisis. Tiny opera, huge invoice.
Timeline
- 2026-06-29 | Kevin articulated the decision rule: there are few one-way doors, and many decisions can be walked back. Source: User, 2026-06-29