Decision Making

There's an interesting discussion going on at Hacker News about using Reason vs. Intuition - but one of the answers to it is quite pertinent to any kind of decision making. I found it worth reproducing for the benefit of all of you to be used in any situation.

There are three tools for discovering truth and making decisions: reason, intuition, and revelation.

Revelation means ask an expert, read the documentation. It is most appropriate when you don't know what you're doing at all -- when you have no sound first principles to feed into the engine of reason, and no experience on which to build intuition. Revelation is fast but limited; you instantly gain a conclusion as sound as your expert, but you cannot improve upon or critique it.

Reason is most appropriate when you have moderate experience in a field. Through revelation and limited experience, you have developed some sound, inviolable principles, and can reason your ways to new ones. You know what must go here because you know what must go there, and you can figure out howthis works because you know how that works. Reason converges to truth slowly but inexorably. It will eventually get there. But it takes a long time to process a lot of input or to navigate a complex landscape.

Intuition is an appropriate tool when you have high experience in a field. Even presented with a complex problem, you know what to do. You just know that OO is the wrong paradigm for this, and that code must go on the server, and this bug is almost certainly caused by a mismatched type somewhere in the parser. Given time, you could probably justify it. Doing so would be equivalent work to writing a paper -- there's just a lot of stuff to consider and weigh. But the power of intuition is that you can decide nearly instantly, and often decide right. Intuition as a tool doesn't converge to truth; it quickly leaps, and then it either gets there or it doesn't. And even when does get there, you don't know immediately whether it actually worked or not. It always uncertain, being only as good as your necessarily incomplete mental models. But there are times when a fast guess is way better than a slow conclusion. And there are times certainty isn't possible anyway. One problem with intuition is that you can't improve its results for a particular problem. If it's wrong, it's just plain wrong. You can't doggedly grind on like you could with reason, or ask another source like you can with revelation.

The highest level of competence comes from using all three tools in concert. With a lot of experience, revelation is sharp: you know who the experts are and what they are likely to know. Reason is sharp: you know a lot of useful rules, and the fastest ways to check truth for certainty. Intuition is sharp: you have a feel for the rhythm of the hills and valleys in the problem landscape. You can guess the existence of a distant mountain the hill-climbing rationalist would take forever to find, quickly and soundly check the local landscape for its unexpected little pits and spikes, and borrow a good map to cross a desert quickly.

This engine of canny guesses and rapid checks and good borrowed maps is precisely what makes communities of hackers work so impossibly, frighteningly fast.

PS: Many thanks to Kalpesh Khivasara (http://twitter.com/kalpeshk) for sharing the link on twitter

PS: Arthavyavastha series will resume in a weeks time. Next Part: Vinimay Avsanrachana (विनिमय अवसंरचना).