The other day I was chatting with some folks at Berkman and this idea came up as a possibly fun and quirky experimental project. It’s not fully fleshed out, but maybe this can inspire more conversations to turn it into something feasible.
The goal of the project is to crate an Internet contraption that, in the spirit of a Rube Goldberg machine, runs on its own after having started it. It will be completely funded, defined, and carried out by the Internet using today’s micro-funding and micro-tasks platforms. The project would be the embodiment of crowdsourcing — it will both celebrate it and problematize it.
The creators of the contraption, will have very little say in the final project itself. Their role would be to define its stages and to set the structure that will hopefully lead to something awesome. Here is how it might work:
Step 1. The instigators will submit the project to a crowd-funding website like Kickstarter with a description of steps 2 and 3. The Kickstarter’s goal will be to raise the necessary funds to pay for stage 2 and 3, and to incentivize the micro-donors, the project will offer more decision power to those who donate more.
Step 2. A portion of the money raised in Step 1 will be used to hire micro-task workers on a platform like Amazon Mechanical Turk to come up for ideas for the final Step 3. The mot interesting ideas would receive higher payment. The micro-donors will get to decide what are the most creative ideas and will let a the crowd decide the one that will become for Step 3.
Step 3. The rest of the money will be used to hire people in the physical world who would carry out the project. These people will be hired on a system like TaskRabbit. Some of the ideas would be to create flash mobs that film themselves and stream the performance live on the Internet.
Some of the obvious questions are:
a) How do you make sure the turkers do not pick something boring, illegal or imposible.
b) What would happen if Step 3 asks people to create another identical project, hence leading to an infinte loop of human computation.
