Getting people to do boring things for fun and profit

September 2, 2006 • 3 min read


This post is over ten years old. Chances are, I've learned enough to have advanced my thinking about some of this stuff.

A bored girl.Much of the world’s money is spent getting people to do boring things. Truck driving, accounting, and toothpaste-tube-screwing-oning all need to be done, and people aren’t exactly lining up to volunteer. While some tasks are being automated by computers (I hear that they’ve almost got verbal communication figured out, those wily researchers) a lot of things are still easier for humans to do themselves.

Coincidentally, a thing has been recently invented called “the internet”. This is a system where millions (if not billions) of bored people go to visit MySpace and read blogs and do other inane things. It seems clear that if these people are already bored enough to be doing these things, even a slight motivation should be enough to get them doing the boring things you’d like them to be doing.

Amazon’s Mechanical Turk

Amazon Mechanical Turk

First up, The Turk. This is Amazon’s service where they pay you insanely small amounts of money to do insanely boring activities. For example, if I’m willing to fill out a 15-question survey about my demographics (including some short-answer questions) I get a whole cent! Alternatively, I’ve been offered 50 cents to scan a supermarket receipt as well as create an excel spreadsheet with all the information about what I bought (including the UPC codes) and send that in. Right.

While Amazon definitely has something interesting here, it suffers from some problems, including the fact that paying somebody to do something makes them think it’s even more unpleasant than they did before. Good luck using this for any large-scale, long-term problem - would you translate a textbook for $0.05 a page? Even worse, if you want a random, intelligent sample of people to do your work then you’re in serious trouble.

The Hollywood Stock Exchange

Hollywood Stock Exchange

HSX is a popular game where you buy and sell stock in how movies will do at the box office. If a movie does well, you earn fame and fortune. If you actually thought Gigli would break even, you’d have gotten a nice cold slap in the face.

So what does this have to do with anything? It turns out that the whole thing is designed to predict what movies people want to go see. People tend to invest in the movies they like the sound of, which works well enough that they make a living selling their research to film companies and theatres.

Google Image Labeler

Google Image Labeler

But what if you have a task more mundane than deciding which movies you think are worth making or seeing? Say, for example, figuring out wtf all those millions of images are of that are choking the internet’s tubes?

Well the folks at Google and Carnegie Mellon University have figured out how to make that into a game too. I wasted way too much time today playing the Google Image Labeler Game (link via Warped Visions). It’s pretty impressive that they’ve made a game that’s both engaging, and more importantly will lead to correctly categorized images. It’s nothing groundbreaking or riveting or anything like that, but is a great example of using people’s play time to get work done for you.

The future

This is just the beginning. As time goes on, we’re going to see more and more of this sort of thing. Maybe the next version of SimCity will feed into a real city’s urban planning database, or maybe the Department of Homeland Security will launch a game where you’re fed routine surveillance videos and report “suspicious activity”. The new generation is bored, wired, and addicted to games, and whether we notice it or not, I think we’re going to be working a virtual job or two on the side.


Liked this? Follow along to see what's next.

© Allen Pike. 👋🏼 Feel free to contact me.