Open Platforms, Legal Stupidity, and Evolution

Mike Linksvayer makes a good point about my triumphalist post last week concerning AOL versus the Internet:

How hard is it to imagine a world in which closed services like AOL remain competitive, or even dominant, leaving the open web to hobbyists and researchers?

One or two copyright-related alternative outcomes could have put open networks at a serious disadvantage.

First, it could have been decided that indexing the web (which requires making and storing copies of content) requires explicit permission. This may have stunted web search, which is critical for using the open web. Many sites would not have granted permission to index if explicit permission were required. Their lawyers would have advised them to not give away valuable intellectual property. A search engine may have had to negotiate deals with hundreds, then thousands (I doubt in this scenario there would ever be millions) of websites, constituting a huge barrier to entry. Google? Never happened. You’re stuck with Lycos.

Second, linking policies could have been held to legally constrain linking, or worse, linking could have been held to require explicit permission. Metcalfe’s law? Never mentioned in the context of the (stunted) web.

He’s certainly right that these things would have been bad, and that bad intellectual property policies in the last decade have probably stunted technological progress in ways analogous to the alternative history he lays out above. But I actually don’t think it’s a coincidence that the web played out the way it did. The Internet’s open architecture was a consequence of having incubated in an open culture. The men (and a few women) who built the Internet did so with the conscious intention of building a decentralized, open network. Along with the TCP/IP protocol stack, HTML, HTTP, SMTP, etc, they also developed a set of norms that emphasized openness and collaboration. The result was that when newcomers like Netscape and Microsoft came on the scene, their efforts to transform the Web into a private fiefdom met with fierce resistance from Internet old-timers.

As a result, by the time some bozo decided to try to assert a right to control what search engines may index, or to decide who may link to his or her content, he was obviously an outlier. The courts could see that, and they left well enough alone. But even if the courts hadn’t deferred to the norms of the Web community, it’s likely the community would have come up with workarounds to the court’s decision. Likely, something akin to a robots.txt file would have been invented that would provide electronic evidence of permission to link, and it would have been bundled by default into Apache. Sure, some commercial web sites would have refused to allow linking, but that would have simply lowered their profile within the web community, the same way the NYT’s columnists have become less prominent post-paywall.

Moreover, I think the right way to look at the Web/AOL/Compuserve story is that commercial services did stunt the growth of an open network, perhaps by as much as a decade. There’s no reason that home users couldn’t have been accessing Usenet and exchanging email in the late 1980s. But they didn’t, because the leading commercial online services firms preferred to keep their services proprietary.

What happened, though, was that the open network, although it started from a smaller base and had less capital behind it, grew faster, thanks (I suspect) largely due to the more powerful network effects that characterize open platforms.

It seems to me that the same thing is likely to happen with other open technologies. In the long run, platforms built on an open foundation grow faster, so sooner or later, the open sectors of the high-tech economy will overtake the closed sectors. Legal restrictions on interoperability don’t stop the growth of open systems, they just force them to start from the ground floor, rather than being built atop previously-closed platforms that had been opened up by third parties.

Had IBM been able to stop the cloning of PCs, (and had IBM or Apple not voluntarily opened their platforms) I suspect we would have eventually seen the rise of another open computer platform. Most likely that would have set back the development of the computer industry by several years, as the commoditization of the PC would have happened several years later. But it would have happened eventually.

By the same token, I think that open music services like eMusic, or perhaps totally free (legal) music communities analogous today’s blogosphere, will eventually undermine today’s DRM-encumbered music marketplace. eMusic is already the #2 music download service, and it’s growing rapidly. Now, it would have been a lot better if MP3.com had won its lawsuit in 2000, because then it’s likely the iTunes Music Store would never have had DRM in the first place. But I don’t think we’re stuck with DRM in perpetuity. Our lousy copyright laws have merely forced an open music market to be built from the ground up, which will take more time.

There’s an obvious evolutionary analogy here: evolution works by trying lots of different variations and then keeping the ones that work. One of the things that made an organism evolutionarily successful was the extent to which its reproductive mechanisms were suited to recombination, experimentation, and selection of beneficial mutations. Sexual reproduction is obviously one such mechanism, as it allows “features” from different individuals within a given species to be combined in new ways with each generation. Another is what Richard Dawkins called “Kaleidoscopic embryology,” which allows gene mutations to produce useful changes (such as a longer or shorter neck) rather than arbitrary, probably pointless ones (such as a random lump of tissue on one side of the neck).

In the long run, organisms that had reproductive machinery well-adapted to rapid, useful modifications tended to win the evolutionary race, because even if they weren’t especially well-adapted initially, their progeny adapted more rapidly to new selection pressures. Even if the organism started out as a marginal species with only a few members and poor adaptation to the initial environment, if they could survive long enough to reproduce for a few generations, we should expect them to gradually displace other organisms whose reproductive machinery was less flexible. The first organism that had sexual reproduction probably didn’t look greatly superior to the other animals around it, but its advantages would have become clear over time, as the descendants of that organism adapted to new environments more rapidly than other, asexually reproducing organisms.

The same phenomenon can be seen in the technology industry. Closed platforms start out with the enormous advantage that they often have billion-dollar companies behind them. As a result, they tend to capture the lion’s share of the market in the early rounds of competition. However, software developed on the bureaucratic model lacks the flexibility and potential for experimentation and recombination of open software, so over time, we can expect the open software to improve at a faster rate. Sooner or later, a tipping point will be reached where the open platform is a better alternative for a significant fraction of users, and users start defecting. Once that happens, network effects take over and as users flee to the open alternative, it becomes ever more attractive to those left in the closed platform.

So I stand by the words “relentless” and “inevitable” to describe the triumph of open over closed systems. I’ll add the concession that the process sometimes takes a while (and obviously, this makes my claim non-falsifiable, since I can always say it hasn’t happened yet), but I think legal restrictions just slow down the growth of open platforms, they don’t change the ultimate outcome.

SHARE: