Should you marry her? - By Ajeya Cotra
TL;DR
Marriage is a contract enforcement mechanism: Cotra argues the point of a wedding is not romance but making it harder to leave for frivolous reasons, while shifting your community toward helping you make the relationship work.
The main upside is investment, not insurance: When both people expect to stay, they put more into shared assets like a house, kids, household systems, and emotional trust, which she says can create a compounding benefit of 10 percent or more over a 10-year marriage.
Waiting only makes sense if you know what you are trying to learn: Using her rough model of one new unit of information per doubling of time together, she says a couple who has dated intensely for 2.5 years may already have most of the information they are going to get.
Judge a spouse in three roles: co-founder, friend, lover: Her checklist ranges from "Would you hire her or work for her?" to "How much would you hang out if it were platonic?" to "How attracted were you in the first 20 minutes?"
Compare her to the actual market, not a fantasy one: Cotra tells confused men to list real women, ask friends for names, estimate how often women of that caliber become available, and honestly ask whether those women are out of your league.
AI timelines change the math: Because she expects a marriage starting today may last only about 10 years in expectation, she says spending 3 years searching means burning roughly 10 percent of the rest of your life, so a replacement partner must be better by a "healthy margin" to justify it.
The Breakdown
Ajeya Cotra frames marriage as a commitment device that pays off through compounding investment, then makes a blunt AI-timelines claim: if marriage might only last around 10 expected years before the world changes, waiting 3 years for someone marginally better is a much bigger cost than most people admit. Her practical test is simple and unusually concrete: assess your partner as co-founder, friend, and lover, then ask whether anyone you could realistically meet would beat her by a wide enough margin to justify the delay.
Was This Useful?
Share
Keep Reading
Make Alcreon Yours
Tune your feedFive quick questions, and the feed ranks what matters to you first.Or just get notified
The weekly Echo. Signal worth keeping in your inbox.
Every new piece, announced on X.
Read Next
See all
Playbook
Tasteful Skills
“Tasteful Skills” argues that the best agent skills are not documentation or best-practice lists.

Playbook
The Art of Tasteful Prompting
Learn how tasteful prompting helps you move beyond generic AI output by shaping context, style, and judgment from the start.

Playbook
The Codex /goal Playbook
OpenAI shipped /goal for the Codex CLI. It turns a prompt into a persisted, self-continuing contract.