CTO Blog

CTO Blog

Opinions expressed on this blog reflect the writer’s views and not the position of the Capgemini Group

Eyes of the Crowd

Category : Open Source

What exactly can be achieved through the unbridled power of crowdsourcing? A nice place to find out is the article on this concept on Wikipedia, which - by the way - itself is created through a mild form of crowdsourcing (these are the moments that you feel that good old, recursive programmers heartbeat again). Luckily enough, Wikipedia points us rigidly to potential terminology errors: crowdsourcing is definitely not the same as open source development and the dynamics of a big mass of anonymous contributors cannot be compared to these of a dedicated community of focused individuals. In any case, it is the Power of Many – enabled by the Web 2.0 platform – that makes results and innovations possible that we never thought of before. The question then is, to what extent. When does the law of large numbers turn against us, when do we find that more collaboration simply does not add value any longer? We know that the quality of software improves with more developers reviewing it. But does that also pertain to knowledge, a proposal or a book? Or even to a transformation program? About software, that’s all described in the classic books of our time. Anybody who want to understand more of the secrets behind the open source success, already sleeps with Eric Raymond’s ‘The Cathedral and the Bazaar’ under the pillow (here’s the article that created the fundament for the book). The subtitle ‘Musings on Linux and Open Source by an accidental revolutionary’ nicely describes what the book is about, but does no justice to the capabilities of Raymond, who is not only a skilled systems developer but also an observant writer. His 19 lessons on why things work or fail in the open source community, are all still unbeaten illustrations of this typical, Bazaar-style of software development. Take for example lesson 8: ‘Given Enough Eyeballs, All Bugs Are Shallow’. To put it differently: given a large enough base of eager testers and co-developers, almost every problem in software will be quickly found and characterized and then understood and solved. There will always be somebody who rapidly finds the problem. And – most likely – somebody else will understand it and take care of it, just as rapidly. Nothing special, according to sociologists. They discovered already many years ago that the averaged opinion of a crowd of equally expert observers (or equally ignorant observers, for that matter) is clearly more reliable a predictor than that of a single randomly-chosen observer from that crowd. They called it ‘the Delphi effect’, which is presumed to be the inspiration to the Delphi method, a formalized approach to creating forecasts or other joint results from a collaborating mass of experts. So software improves with more developers and testers involved. It sure looks like Fred Brooks’ Mythical Man Month needs a slight revision after all. On the other hand, when Brooks wrote his famous book (long before Raymond did), nobody saw Internet even remotely coming. It’s nothing less than a different view of the world. If collective wisdom works just like that in the real world, outside of software and systems, is yet to be determined. It requires a lot of unraveling and simply experiencing it in practice. We already established that Wikipedia is a convincing example of what a crowd can achieve: given enough eyeballs, the articles of Wikipedia become more and more up-to-date, complete and consistent. That is mandatory inspiration to anybody who wants to revitalize knowledge management within an organization. Even so, Amazon’s Mechanical Turk proves that you can always find somebody to execute a hard task or answer a difficult question. As long as a motivated crowd is connected in sufficient numbers. The fact that this artificial artificial intelligence now mainly focuses on, well, basic issues (‘find a picture with purple clouds’, ‘who is playing bass guitar in this song’) seems just a matter of inexperience. We can only speculate on what will happen if we combine the joint expertise of say 65.000 business and IT consultants in exactly the same way to answer a single, complex question of a client in the shortest conceivable time. With perfect results. Knowing all this, are there any areas where we simply would have too much of a good thing? I think this deserves a separate blog-item, soon from now. It will allow me to elaborate on phenomena such as mass auditions, collective proposals and even transformation crowds. Also, I will introduce you to an interesting project in which a million penguins are writing a book about Fluffy the cat and lots and lots of bananas. Don’t ask. For now.

About the author

Ron Tolido

Leave a comment

Your email address will not be published. Required fields are marked *.