In a recent article, renowned virtual reality guru Jaron Lanier criticizes the community-driven software development process of the open-source movement, asserting that the model isn't conducive to radical innovation. Lanier believes that the scientific community should shun the open approach and not follow in the footsteps of Linux.
Lanier is a noted critic of Internet collectivism, and one can look at his previously stated positions in order to gain a better understanding of his opinions about open-source software. In a 2006 essay titled "Digital Maoism," Lanier condemned Wikipedia as a "fetish site for foolish collectivism" and cautioned against what he perceived as descent into mob mentality. The basis for Lanier's concerns about Internet collectivism, which he elucidated in an article for TIME magazine later that year, seem to extend from the observation that collectivization has a dehumanizing effect that erodes moral autonomy and reduces individuals to the lowest common denominator of the group.
Like most people who have had close encounters with idiocy on the 'Net, I can certainly sympathize with many of his concerns. In the article, Lanier points out that anonymity tends to foster abusive attitudes and that a collective of anonymous people can easily become a brutish gang. The correlation between anonymity and abrasiveness, especially in the presence of an audience, is quite apparent on the Internet (we may even be treated to a demonstration in the discussion thread for this article). Lanier doesn't directly address the issue of how collectives magnify the effect, but I think the reasons are pretty obvious: in a collective, participants mutually reinforce each others' behaviors.
The open-source software community certainly exhibits some of the negative traits of collectivism. For instance, there is a very strong tendency to erect personality cults around strong leaders whose opinions are then given undue weight. There are also many cases where certain groups or factions of the open-source software community have resorted to nasty personal attacks and other inappropriate mob-like behavior in response to a perceived threat. I can also recall several notable instances where irrational group consensus led to adoption of poor practices and technologies to the detriment of innovation. One example of this is CORBA, which was adopted with way too much enthusiasm on the Linux platform despite serious deficiencies and has now been almost completely purged.
The upside to community-driven development
Although there is evidence of counterproductive groupthink in the open-source community, that doesn't really invalidate the viability of the community-driven development model. Obviously, corporations that make closed-source software are also collectives and are just as susceptible to the same exact negative tendencies. Universities and the scientific community in and of itself also represent certain kinds of collectives. I don't think it is possible to operate or innovate in complete isolation. Conventional development and community-driven development both ultimately have to entail collectives of some kind, the difference is just that community-driven development is infinitely more inclusive and transparent.
I was somewhat puzzled by some of the examples and anecdotes that Lanier uses to support his claim that open source software is detrimental to innovation. Lanier describes how conflicts with Symbolics over the source code relating to MIT LISP machine development inspired Richard Stallman, the founder of GNU, to pursue development of an open-source operating system. The general gist of the story is known to most open-source enthusiasts, but Lanier's interpretation is somewhat unconventional. "[Richard] would instigate a free version of an ascendant, if rather dull, program: the Unix operating system," Lanier wrote in his article. "I was intrigued but sad. I thought that code was important in more ways than politics can ever be. If politically correct code was going to amount to endless replays of dull stuff like Unix instead of bold projects like the LISP Machine, what was the point?"
The point, which Lanier apparently misses, was to create technology that was practical and useful while also empowering users by granting them the freedom to study, modify, and redistribute source code. The ability to freely reuse existing components promotes innovation by enabling independent developers to focus on adding new features and experimenting with new ideas rather than having to start from the ground floor or simply accept the limitations inherent in software that they cannot modify.
Lanier's attempt to cite the cloning of UNIX as an example of open-source anti-innovation is absurd because commercial, closed-source software companies were creating proprietary UNIX clones at the same exact time. Lanier is trying to build an argument against open source on the basis of something that isn't exclusive to open source. Although only tangentially related, I think it is also worth noting that the LISP Machine failed, in part, because it didn't provide many technical advantages over mainstream UNIX workstations that offered superior software portability. UNIX was a practical platform then, and the modern incarnations are practical platforms now.
True innovation from open source
I was also somewhat puzzled by the specific examples of closed-source innovation that Lanier chose to defend his argument. He considers products like Adobe Flash and the Apple iPhone to be paragons of proprietary innovation. Lanier fails to note that many important components in both of those products are built with open-source software. Adobe opened the source code of its Flash ECMAScript implementation and at the heart of the iPhone is Apple's Mac OS X operating system which leverages BSD and a kernel derived from the open source Mach kernel.
Using Mach and BSD for Mac OS X, and then using Mac OS X as the basis for the iPhone's OS made it easier for Apple to focus on developing the user interface and other features that make the iPhone innovative. In that respect, open-source software serves as a compelling enabler of Apple's innovation by providing a robust and flexible foundation that Apple could adapt to serve its purposes.
Many companies are building innovative new technologies on top of a foundation of open source software. TiVo and the Amazon Kindle, for instance, are both examples of innovative proprietary products that are Linux-based, while the Chumby and the Neo1973 are examples of innovative open-source products that are Linux-based.
In the end, Lanier really doesn't provide any compelling evidence to support his claim that adoption of community-driven processes hinders creativity. His fear that innovation in the scientific community will stagnate as a result of adopting open-source methodologies seems largely unfounded. There is a growing body of evidence—based on actual studies rather than mere speculation—that the community-driven model is, in reality, more productive and more conducive to innovation.
reader comments