Crowdsourcing has become a powerful mechanism for accomplishing work
online. Hundreds of thousands of volunteers have completed tasks including
classifying craters on planetary surfaces, deciphering scanned text, and
discovering new galaxies. Until now crowdsourcing has worked especially
well for tasks that are fast to complete, incur low cognitive load, have
low barriers to entry, are objective and verifiable, require little
expertise, and can be broken up into independent subtasks. However, these
simple tasks tap into only the lowest levels of human intelligence. In
this talk I discuss ways to accomplish creative and complex work through
crowdsourcing, including poetry translation, article writing, scientific
journalism, and information seeking. I will present the science and
systems we have developed to support task decomposition, quality control
and coordination in subjective, complex, and creative crowd work. This
work aims towards a future which realizes the potential of crowdsourcing
for science, education, and industry to enable better decisions, greater
innovation, and more literate, knowledgeable and intelligent citizens.
