The Weekly Better Rule
End every week better than you began it, on any axis you choose
- Difficulty
- Easy
- Time to result
- ~ongoing to results
- Steps
- 5
- Confidence
- 80%
Vitalik Buterin's answer to 'what philosophy helps you make decisions' is a single recurring test rather than a life plan: strive to be a better person at the end of every week than you were at the beginning of it. The mechanism is a weekly boundary plus a deliberately loose definition of 'better' — learn something, get fitter, improve how you interact with people, or add a genuinely new experience such as visiting a country you have never seen. Input is one intention per week; process is living the week; output is a one-sentence verdict at the boundary. Because the definition is broad, the rule almost never fails on a technicality, which is what makes the cadence survivable. Naval's closing framing supplies the compounding logic: improvement is a function of the number of iterations, not of elapsed time, so shortening the loop to a week is the whole lever. The failure state the rule guards against is explicit — living life on autopilot.
Origin
Given by Vitalik Buterin as his closing advice to young listeners on the Naval podcast, in answer to whether any conscious philosophy guided his own unconventional path off the standard school-degree-job track.
Core principles
- 01The week, not the year, is the right unit of self-improvement.
- 02"Better" is deliberately broad — knowledge, fitness, relationships, or novel experience all count.
- 03Progress compounds with the number of iterations, not with elapsed time.
- 04A week you cannot point to any improvement in is a week spent on autopilot.
- 05You do not need a grand plan; you need a cadence you can never fall behind on for long.
How to run it
- 1
Fix a weekly boundary
Choose one recurring moment each week that marks the start and end of the cycle. The boundary must be non-negotiable, because the whole framework is the cadence rather than the content.
Pro tip Attach it to something you already do weekly so it survives travel and disruption.
- 2
Name the axis for the week
At the start of the week, decide what 'better' will mean this time: learning something, physical improvement, improving how you interact with other people, or adding a new experience. One axis is enough.
Pro tip Rotate axes deliberately so one dimension of life does not quietly absorb every week.
Watch out Naming three axes at once usually produces zero, because none of them has a clear verdict.
- 3
Make the smallest real move on that axis
Do something concrete enough that you could describe it to someone else. Getting better at running than you were before, or visiting a country you have not visited, both qualify; intending to do so does not.
Watch out Do not scale the move to what an ideal week could hold; scale it to what a bad week can still deliver.
- 4
Deliver the one-sentence verdict
At the boundary, state how you are better than you were seven days ago. If you cannot produce a sentence, the week is recorded as wasted rather than argued into a pass.
Pro tip Keep the sentences in one running file — the list itself becomes the evidence that the loop works.
Watch out Retroactively redefining 'better' to make the week pass destroys the signal the rule exists to produce.
- 5
Reset and iterate immediately
Do not carry a failed week forward as debt or motivation. Start the next cycle clean, because the value comes from the number of iterations completed rather than from any single week's quality.
Pro tip Track iteration count, not streak length — a broken streak tempts you to abandon the practice entirely.
In the wild
Buterin left the standard track early and has been a nomad for roughly eight years, describing his life over one six-month stretch as San Francisco, Toronto, Argentina, Costa Rica, Mexico and Singapore. Rather than following a career ladder, he accumulated languages — somewhere between four and six, with Chinese picked up over five years — and kept adding new places and new problems. The pattern matches the rule he articulates: each interval adds a discrete increment on some axis, without a master plan tying them together.
→ A self-directed path with no degree-and-job scaffolding that nonetheless compounded into deep language ability, a global network and a multi-act technical career.
An engineer adopts the rule with a four-week rotation. Week one is learning: finish one dense paper and write a paragraph summary. Week two is physical: add half a kilometre to a weekly run. Week three is social: repair one relationship that has gone quiet. Week four is novelty: spend a day somewhere in their own city they have never been. Each Sunday they write a single sentence in a running file. After three months the file has twelve sentences and two of the axes have clearly become self-sustaining habits.
→ Twelve logged iterations rather than one abandoned annual resolution, with two habits that no longer need the prompt.
Common mistakes
Defining 'better' too narrowly
Restricting the rule to career or productivity gains makes most weeks fail, which kills the cadence. Buterin's definition deliberately includes exercise, social ability and simply visiting somewhere new.
Optimising the week instead of the loop
People try to make each week impressive rather than making the review reliable. The compounding comes from the number of iterations, not the peak quality of any single one.
Skipping the verdict when the week went badly
Avoiding the review after a poor week is exactly how autopilot returns. The honest 'this week was wasted' entry is what makes the next iteration deliberate.
Is it for you?
Best for
Someone early in their career who wants a low-overhead, self-directed alternative to the school-degree-job default path.
Not ideal for
Anyone who needs an externally-assessed, milestone-based programme with formal accountability and deadlines.
From the transcript
“strive to be a better person at the end of every week than you were at the beginning of it”
“keep on moving forward and you'll be fine don't live life on autopilot”
“compound interest is a function of number of iterations not time”
From the episode
Vitalik: Ethereum, Part 2