Academic journal article Research-Technology Management

The Limits of Crowds

Academic journal article Research-Technology Management

The Limits of Crowds

Article excerpt

As of this writing, and after 88 days of concerted effort, the oil spill at the site of the Deepwater Horizon drilling rig in the Gulf of Mexico has been capped, at least temporarily stopping an unprecedented flow of oil into the ocean. Given the damage done to date, there was little celebration and there will be many recriminations.

The Gulf oil spill is the first major technical disaster in the age of social media, and social media have shaped our response to the disaster in several ways. The oil itself is constantly visible, available for anyone to monitor via a webcam at the mouth of the well. And the blogosphere has provided a flow of information and opinion-about BP, about the energy economy, about nature conservation--at times as powerful as the flow of oil.

The response to the disaster has also been shaped by the use of social media, as BP and other interested parties have undertaken several initiatives to leverage the expertise of the general public. These have included a wiki, challenges sponsored by InnoCentive, and sites like WhatShouldBPDo.com. The U.S. government's Deepwater Horizon Unified Command launched a suggestion site that reportedly garnered 20,000 ideas. And various tinkerers, inventors, and entrepreneurs have posted videos on YouTube proposing ways to plug the well or clean up the spill.

These initiatives were broadly applauded. There were stories in the media and blog entries about the innovative ideas of individuals and about BP's resistance to evaluating them, including commentary from visionaries like Don Tapscott, co-author of Wikinomics. Underlying the praise for these efforts is a new set of assumptions about the nature and value of expertise, influenced by the rise of social media and especially by what has come to be called "crowdsourcing." Our trust in the expert appears to be increasingly supplanted by a willingness to rely on the knowledge derived from crowds of amateurs. In this new world, the motives and competence of experts are at best suspect and presumed to be inferior to the wisdom of crowds.

The general approval surrounding crowdsourcing has included little discussion of its limits and the challenges of making it work in practice, on real, complex, ill-defined problems. Crowdsourcing is a transformative capability, with wide application in fields ranging from software development to parasail design. But it will not work for all problems and it will not work without a concerted, planned effort.

Effective crowdsourcing requires a number of conditions:

1. The problem (and its boundary conditions) must be well defined,

2. The population of potential solvers with relevant expertise must be large,

3. Feedback must be provided to the crowd (not just to individual contributors) so that ideas can evolve,

4. Mechanisms for managing intellectual property must be in place, and

5. Someone needs to filter the ideas (and develop them).

For the Deepwater Horizon well, the problem can be stated simply ("stop the flow of oil"), but the boundary conditions are manifold and complex. What is the tubing made of? What does it take to cut through or drill into the tubing? What are the actual flow rates and pressures? What chemical reactions might interfere? What are the capabilities of the robotic vehicles? What might the consequences of plugging the well be below the surface? Any solution must be cognizant of these and many other factors. Solutions that work on paper might have significant difficulties in practice.

There are, in this instance, many problem solvers willing to contribute their ideas, so at first blush it may seem that the second criterion is satisfied. A large percentage of the contributors, however, have limited relevant expertise. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.