Май 2019 — Заметка №4

Slava Akhmechet (был продактом в Stripe, построил RethinkDB) пишет в Твиттере (с тех пор он удалил эти два своих твита) описания стратегий, которые помогают решать сложные проблемы. За несколько лет наблюдений за кучей проблем и кейсов он обнаружил, что их решения сводятся к ~20 стратегиям. Пока он запостил две.

1. Radical quantification.

TL;DR: transform an arbitrarily complex problem into a trivial one by picking just one quantifiable dimension to worry about.

Свести проблему к одному главному числу и думать только о нем.

Here is a trivial example: Net Promoter Score. You take an extremely complex problem of product/service satisfaction and distill it to one number – “would you recommend this to others?”

Consider a problem: how much did real wages grow from 1800-1992? The problem is extremely complex because it requires properly computing inflation in presence of rapid technological change and incomplete economic data. Is there a way to solve it with radical quantification? Here is a solution by William Nordhaus (Nobel Prize in economics in 2018). Reduce the problem to a single dimension – how many hours did the average person have to work to produce a lumen of light? This isn’t easy to compute, but it makes an impossible problem tractable.

You can project nearly every dimension into another one - money. That lets you compare seemingly incomparable things.

One of the coolest things about radical quantification is that you don’t even have to pick a dimension you can observe because the technique is extremely susceptible to using proxy metrics. You just have to be really creative about picking the unit.

2. Near/far transformation.

TL;DR: our brain has two sets of heuristics to quickly parse complex streams of information. Manually tricking it to shift into a different mode can make seemingly intractable problems trivial.

Переключать себя между двумя режимами: ближний (личное, детали, эмоции, внимание) и дальний (асбтрактное, объективное, взгляд сверху, отстранненность).

A way to make a difficult problem simple is to reframe a far mode problem in near mode. Simple example. Far mode: how do I pick a VC? Near mode: can I stand working with Bill for ten years? You’ve outsourced a hard problem to your gut for an immediate answer!

Another similar trick: convert a continuous space into discrete units. Paul Graham did this in his Life is Short essay (http://www.paulgraham.com/vb.html). How many Christmases do you have with your children? Waitbutwhy did it with Your Life in Weeks calendar.

Another example of far->near mode. Here are WhatsApp founders describing the very complex question of online advertising: “no one wakes up excited to see more advertising; no one goes to sleep thinking about the ads they’ll see tomorrow”.

We tend to think of ourselves in near mode (great detail) and of others in far mode (abstractly). Flipping the two can make us seem dramatically more attentive. And when you hear internet advice that “nobody thinks about you that much”, what it really means is that when they do think about you, they think about you in far mode. Of course there is the opposite trick too – trick yourself to shift from near mode to far mode to get some objectivity.

Describing current events in far mode can make one seem more profound by giving people some objectivity. Shifting your mind into far mode can help you make objective political decisions (especially if you’re highly attached to your current reality, which we usually are). “So if you’re considering the worthiness of a particular statue […] Pretend you’re from some very distant foreign country and view the dispute through that more objective lens.”

A far mode trick that can help make decisions – imagine multiple futures and pick between them. […] And a near mode version of this – imagine you’re already in a counterfactual future, and decide whether you’d want to come back.