Why Zittrain's Techno-Pessimism Is Unwarranted
from the no-worries dept
Ars Technica reviews Jonathan Zittrain’s new book, The Future of the Internet and How to Stop It. Zittrain is by all accounts a smart guy and an engaging speaker, and it sounds like his book makes a lot of worthwhile points about the importance of open, “generative” technologies. But I just can’t get worked up about Zittrain’s warnings that the dominance of open systems is a fragile, temporary thing. It seems to me that there’s a basic tension at the heart of Zittrain’s argument. On the one hand, he argues (correctly in my view) that open platforms are better for innovation because of their lower barriers to entry. On the other hand, he wants us to believe that despite that inherent advantage, open technologies are on the brink of being eclipsed by closed platforms like the iPhone.
I think this misses a couple of important points. In the first place, I think Zittrain draws the wrong lessons from history. Zittrain himself notes that until the 1990s, the world was full of proprietary networking technologies and computing platforms that had big advantages over open technologies like TCP/IP, Unix, and the mostly-open PC platform. Open technologies had a few advantages of their own — most notably government support of TCP/IP — but open platforms were definitely the underdogs in many respects. And then, of course, the open platforms utterly destroyed the closed ones. Almost everyone now uses TCP/IP, while AOL is now little more than a mediocre website. Virtually all desktops and laptops — including Macs and a lot of Unix workstations — now largely share a common architecture. And almost every operating system not made by Microsoft is built on some versian of Unix.
Zittrain would have us regard all of this as some kind of fluke or lucky break, that the whole thing could come crashing down at any minute. But I think it’s evidence that better technologies tend to win out in the marketplace. TCP/IP beat out AOL and other proprietary services precisely because open architectures enable more innovation. And once an open architecture comes to dominate a given market, it becomes harder, not easier for a proprietary product to displace it, because network effects create tremendous intertia on behalf of established open standards. I’m hard pressed to come up with any examples of a well-established open standard getting displaced by a closed one. Rather, what tends to happen is that new, proprietary technologies tend to get built on top of open ones. The top layers of the iPhone software stack may be closed, but it’s built on TCP/IP, HTTP, and a host of other open standards.
It doesn’t, therefore, make sense to view the iPhone as a threat to “generativity.” The iPhone itself may not be “generative,” but it’s built on the same open standards as more open devices. That means that growing the iPhone market is a net positive for openness overall. True, people who buy an actual iPhone aren’t getting the full advantage of generativity, but they are helping to further entrench TCP/IP and the web, platforms on which other more generative technologies can thrive alongside the iPhone. Moreover, if Zittrain is right that open platforms promote more innovation, which I think he is, then we should expect the same thing to happen at the top of the stack as happened at lower layers of the stack: over time, open mobile platforms like Android should enjoy more innovation than closed platforms like the iPhone, and the former should gradually displace the latter. Consumers tend to choose more open platforms over time not because consumers care about “generativity,” per se, but because they want the phone with the best software, and open platforms tend to get the best software over time. And smart companies will tend to open up their platforms over time, lest competitors leapfrog them with a more open product. Indeed, as Mike pointed out a few days ago, that’s already happening with Nokia’s decision to open source its Symbian operating system.
Filed Under: closed, jonathan zittrain, open
Comments on “Why Zittrain's Techno-Pessimism Is Unwarranted”
Open systems do more than just encourage innovation
While it’s true that open platforms do encourage more innovation in their space it’s also true, given the realities of things like TCP/IP and the World Wide Web that they foster cooperation between vendors along with competition to produce the best quality products.
W3C is a prime example of this where standards are set for the web in a cooperative environment that enforces that through its policy on patents. (They’re a big no-no for web standards.)
The same is also true of the Linux Standards Base though some would correctly argue that it hasn’t gone far enough yet.
When innovation and competition occurs in an open space both are enhanced both in quality and effectiveness.
Yes, the iPhone is a proprietary shell on top of a collection of open standards. Mac OSX is as well. So, increasingly, is Windows.
The same is true for cultural pursuits only to a much much larger degree.
None of which says that people ought to work for nothing or to not be rewarded for their efforts. Far from it.
It only goes to illustrate that everything is built on something else that came before. Be it a book or an invention.
Being open always trumps being closed.
ttfn
John
Tendency Does Not Equal Certainty
I am working my way through Mr. Zittrain’s book, and I saw him speak at Personal Democracy Forum 2008 (his presentation was recorded and is available here). So while I’m not an expert on Mr. Zittrain’s theories, I do see some holes in your rebuttal.
You are comparing apples and oranges. Open architectures do enable more innovation, but that’s a long term trend, best expressed as some sort of probability. For any given fork in the road between an open and closed rendition of the same technology, the open one may win out. For the sum of all such forks in the road, the open ones will tend to win out. So, in the concrete case of TCP/IP vs. AOL, you can’t just say that it’s because TCP/IP was an open architecture as a fait accompli — you actually need to prove that point with some evidence that, indeed, it was innovation on the TCP/IP-based Internet that caused it to win out. I have no doubt that’s the case, but just because it worked here does not mean it will work 100% of the time. Otherwise, we’d all be running Unix.
Network effects create tremendous inertia on behalf of the winning standards, whether or not they are open. Network effects are not limited solely to open standards. However, since open tends to beat closed, the network effects tend to have less staying power with closed.
I agree that, for comparable standards, closed rarely if ever displaces open. However, closed may well replace open in a new generation of technology, either on a broad basis or in a niche. Plain text (open) used to be the norm for textual documents. Then, word processors were invented, and there was a soup of formats, until Microsoft Word (closed) dominated the tools and the format. Unix (sorta open) used to be the norm in operating systems on minicomputers, but then the microcomputer was introduced, we had a melange of operating systems, until Microsoft Windows (closed) became the standard platform for third-party applications. Programming languages have waxed and waned, many open, some (e.g., Visual Basic) closed.
It’s like saying that the prevailing winds for a region of the ocean tend to be from the west — that doesn’t mean that a schooner will make a beeline due west. Other pressures (currents, storms, morons setting up the sail plan) will have more tactical impact on a given schooner’s course than will the prevailing winds. Similarly, open-beats-closed is a tendency that will play out over time, but other pressures (e.g., laws, oligopolies) may impact any given technology over a short timeframe. I believe Mr. Zittrain’s “techno-pessimism” might be better viewed as a spin on Reagan’s “trust, but verify” line — if you value openness, you can’t assume everything will always be open, so you have to work at it.
Re: Tendency Does Not Equal Certainty
Plain text (open) used to be the norm for textual documents. Then, word processors were invented, and there was a soup of formats, until Microsoft Word (closed) dominated the tools and the format.
I think this rather supports my point, though. Word didn’t replace ASCII. ASCII’s still around and supported by all modern computers. Moreover, the Word format is gradually being transformed into an open standard by Open Office, Apple, and others. Finally, .doc has a lot more competition than it used to. The dominant document formats these days are arguably HTML and PDF, not .doc.
Also, while I dislike Windows as much as the next geek, it was the comparatively open option during the period when it achieved its dominance. Anyone could build a Windows PC and anyone could develop a Windows application. The same could not be said of Microsoft’s principle competitors during the same period. Finally, remember that Unix continues to be widely used in universities and on servers, the same places that it was used in the 1970s. Windows didn’t displace it, it simply conquered a new market where Unix never really got a toehold.
Re: Re: Tendency Does Not Equal Certainty
“Windows didn’t displace it, it simply conquered a new market where Unix never really got a toehold.”
Let’s not forget the fact that proprietary vendors of UNIX at the time were basically selling incompatible versions of the system and wildly inflated prices.
Linux is slowly tipping that though Windows is so entrenched it will take time and/or another horrific mistake like Vista to accelerate the desktop penetration of Linux.
Then again, the tiny laptop market may just do that sooner rather than later given that Linux scales well and Vista not at all. Even the XP sold on those machines isn’t fully the equal of the XP found on full desktops or laptops.
ttfn
John
iPhone a poor example
“…open technologies are on the brink of being eclipsed by closed platforms like the iPhone”
The iPhone is not an example of a closed system replacing a “collapsing” open market because the phone market has always been closed. The iPhone and other recent devices are the start of opening up this market which will be opened up further when the Android phones become available.
Paul
I’ve got to wonder if that still holds true, actually. Back in the 90s, most computer equipment purchases were still coming either from professionals or enthusiasts, who invariably had a fairly significant degree of technical knowledge. An everyday user, who just wants something to browse the Internet and fill their iPod with, is going to have more trouble filtering through the sales pitch to the facts underneath; this is no reflection on them, of course, because they’ve not had the same training or experience as people who have to make computers work as part of their day job. If the iPod can coast along on slick marketing despite its deeply flawed software bundle (or at least the PC port of it; iTunes may be a paragon on OSX for all I know) and extensive lock-in, so can others.
The value of "handholding"
I think you make a lot of good points. One more point along this line. There will always be a marketplace opportunity for vendors to provide tools that have limited functionality because some consumers don’t want power and customizability; they just want the device to do a few things really well in a way they can learn and understand, and they are willing to give up the generativity capacity as a tradeoff. This isn’t a failure of open systems; it’s just a different market segment. Eric.
the iPhone and open source
The only thing the iPhone demonstrates is that not all good ideas are open source.
Also there’s a critical difference between iPhone and most technologies. iPhone is designed to work with a mostly closed system — the phone system. The iPhone equivalent of something like WiFi wouldn’t last long.
The evidence is clear from a dozen fields: When proprietary goes up against open, proprietary loses — all things being equal. (Which they are not with the telephone system.)Consider computer networks as a counter example. Virtually everything is built on open source like Ethernet, even in highly specialized areas like factory automation.
I’m sure some people will find Zittrain’s conclusions comforting. I just don’t find them persuasive.
ahhh?
why such diluted agrument of open vs closed tech, what about the obvious one?
PC (open) = 90+% of worldwide market
MAC (closed) =
Unix Open?? iPhone Closed?? Whoah There!
Okay, first, everyone is throwing Unix around as if Unix were an open platform. Linux and [Open|Free|Net]BSD are of course fine examples of “Open Source” operating systems, but UNIX, back in the 90’s, referred primarily to AT&T UNIX, and it ran on a few platforms, but it was not all that “open”. True, there were a lot of applications which could travel from one hardware platform to another, and there was more openness built into UNIX than Microsoft built into the early versions of windows, but I attribute that more to its age than any real willingness on the part of UNIX vendors to give up any control. If they were open, it was because the scientific and educational markets, where a large number of their systems were deployed, were peopled with the kind of users and admins who demanded access to things, and wanted to fix something when it broke. The BSD’s are based on source that came out of a project at Berkely, and Linux was one guy’s project to make something unix-like run on PC hardware. Neither of them were supported or sponsored by any UNIX vendor directly. So Unix is a poor generic term for the openness which is the Linux phenomenon. Commercial UNIX vendors have opened up much more since the rise of Linux, but that was an effort to compete, not because they wanted openness for their own sake.
And, lest we forget, what OS does the iPhone run? Why, it runs a scaled-down version of OS X, which is, at its core, down below the cocoa UI, a BSD based operating system. So the iPhone does have some openness in its pedigree. One could say that without open software, OS X would not have been possible. Of course, apple is currently operating the iPhone platform as a private preserve, but I think those walls will come down in time. Steve Jobs doesn’t really do “Open”, which is why we don’t see a bunch of commodity PC hardware around running Apple’s OS. Macintosh is not exactly a poster child for openness. They provide the things software developers need to write applications, just like Windows, but the core of the OS is still Apple proprietary.
Re: Unix Open?? iPhone Closed?? Whoah There!
Quick correction: Linux is far more than ‘one guy’s project’. Linus Torvalds wrote the Linux kernel, but the entire rest of the operating system he just found lying around because Richard Stallman and his GNU Foundation had been writing their own free, open unix-like operating system for *years*. If it hadn’t been for the enormous effort Stallman dedicated to the free software community, Linus would have just been another random hacker who wrote an OS kernel.
"Closed" IPhone not so closed
The idea that the Iphone is closed is interesting since a great deal of its popularity is due to the “jailbreaking” of the device and allowing it to be more open. I believe this is a big driver for Nokia’s decision as they saw that even with a “closed” system, the Iphone is based on open standards and once it was opened up, a wealth of simple apps have sprang up and many people are using those apps and driving the Iphones popularity. If the Iphone OS was not Unix/Linux based, I don’t think you would see the vast collection of apps available for “jailbroken” Iphones as it would be much more difficult to port to a proprietary OS and not as many volunteers would have experience coding for that proprietary OS. This I believe provides more evidence that Zitran’s position is flawed.
Time traveling here from 2023 to say: Closed platforms like iPhones and the App Store have indeed caused a lot of harm to the idea of open platforms.