§ 05 Prediction Notes
Three timescales.
Three risks.
Three bets.
The closing section of every issue exists to do something almost no publication is willing to do: place dated, specific, falsifiable bets on the future, and live with the consequences.
The architecture
First
12 months
By an active founder.
The view from the inside of the next twelve months — written by someone whose decisions next quarter depend on getting this right.
Second
3 years
By a Series A–B investor.
The window in which thesis-driven capital is currently being deployed. Specific enough to be checked. Sharp enough to be controversial.
Third
10 years
By a Stanford faculty member.
The horizon at which structural changes — in education, in capital, in technology — resolve into something legible. Borrowed authority on the longest timescale.
The Architectural DecisionThe editorial team never writes the predictions.
Predictive authority cannot be claimed by a young publication. It can only be borrowed and, over time, earned. We borrow it by commissioning each timescale from a different category of person, and we earn it through a mechanism we will run once a year.
FormHow a prediction is written.
- 500 to 1,000 words. Short. Sharp. Dated.
- At least one non-consensus claim per prediction. Soft consensus is rejected.
- Specific enough to be checked against reality on a fixed timeline.
The Compounding MechanismThe Prediction Audit.
Once a year, the magazine publishes a Prediction Audit — a public scorecard of which previous predictions held up, which did not, and which were ambiguous.
Over a decade, the archive itself becomes the asset. A publication that signs its predictions in public, dates them, and grades them honestly is doing something almost no one else is willing to do — and the cumulative record is one of the very few credible historical documents of how this generation of founders actually saw the world.
A Note on DiscomfortWhy this is the hardest section.
Most publications avoid concrete predictions because they make their authors look stupid in retrospect. We have decided that looking stupid in public, where stupidity can be verified, is preferable to never having said anything verifiable in the first place.
This is the section the editorial team will be asked to retract from most often. The discipline is to never retract a prediction silently. Anything wrong is graded wrong, in writing, in the following year’s audit.
A prediction unaccompanied by a date is a wish. A prediction unaccompanied by a falsification condition is a prayer. Stanford Entrepreneurs publishes neither.