The Future of Crowd Work

Can we foresee a future crowd workplace in which we would want our children to participate?

There is a serious risk that without intervention the future of crowd work will continue down the path of diminishing pay rates, replacement of skilled with unskilled labor, and a lack of educational and career development for the workers involved. How can we move towards a future of crowd work that is more attractive for both employers and workers than existing systems? What new services, systems or features are needed for a future of crowd work that the reader would be proud to see his or her children take on as their livelihood, and how can we go about realizing these?

Current crowd work typically consists of small, independent, and homogenous tasks:
crowdwork_current

Workers are paired with an instance of each task to produce an output. Such simple, small-scale work has engendered low-pay, piece rate reward structures, in part due to the perception that workers are homogenous and unskilled. The current model is also insufficient to support the complexity, creativity, and skills that are needed for many kinds of professional work that take place today. Nor can it drive factors that will lead to increased worker satisfaction, such as improved pay, skill development, and complex work structures.

Instead, much professional work consists of complex sets of interdependent tasks that need to be coordinated across individuals with different expertise and capabilities. For example, producing a book, an academic paper, or a new car all may involve many individuals working in structured teams, each with different skills and roles, collaborating on a shared output. How can crowd work support such complex, creative and interdependent work?

To address this we draw on two fields that have solved some of these problems in very different domains: organizational behavior and distributed computing. Both distributed organizations and computing systems face many common challenges in accomplishing complex work, for example partitioning tasks into subtasks that can be done in parallel, mapping tasks to (human or machine) processors, and distributing data to and between processors. Specifically, we develop the beginnings of a framework for the future of crowd work that integrates the human aspects of organizational behavior with the automation and scalability of the distributed computing literature.
crowdwork_future

This framework highlights twelve key research foci we suggest are necessary for realizing a positive future of crowd work, ranging from task decomposition to reputation systems to quality assurance. For each of the foci we introduce its importance, survey existing work, and propose key research challenges that need to be overcome. We then provide three design goals that demonstrate how the integration of multiple foci can lead to concrete next steps and calls to action that will create a better environment for crowd workers, and better work for employers.

Crowd work may take place in minutes, but the impact of crowd work may be felt for generations. We have asked: what will it take for us, the stakeholders in crowd work – including requesters, workers, researchers – to feel proud of our own children when they enter such a work force? We hope the community’s observational, experimental, design and technical skills will play a vital role in shaping the future of crowd work and the next generation of workers.

For more, see our full paper, The Future of Crowd Work, presented at CSCW 2013.

Niki Kittur, Carnegie Mellon University
Jeff Nickerson, Stevens Institute of Technology
Michael Bernstein, Stanford University
Elizabeth Gerber, Northwestern University
Aaron Shaw, Northwestern University
John Zimmerman, Carnegie Mellon University
Matthew Lease, University of Texas at Austin
John Horton, oDesk

One thought on “The Future of Crowd Work

  1. Pingback: TurKit Retrospective: An Interview with Greg Little | Follow the Crowd

Comments are closed.