Flipping through Shine, I ran across an interesting mention (in the introduction to Lavie Tidhar’s contribution) of Jan Romein’s notion of Wet van de remmende voorsprong, or, in English, the “Law of the Handicap of a Head Start.”
It struck me that the Korean internet (a subject about which I recently posted) is a wonderful demonstration of both sides of this notion: the benefit of a late start, and the handicaps that a head start can introduce into a system.
The link above details several of the handicaps: the dependency on Windows, ActiveX, and even Internet Explorer 6.0, as well as the general institutional resistance to retooling Korean net commerce and web design to modern, global standards, multi-platform functionality, and so on. All of this results from the early adoption of a locally-developed system for security that the Korean government developed, dependent on ActiveX controls, so that ecommerce could begin earlier, instead of waiting for American encryption technology to be exported. That is to say, Korea innovated its own software, got into the game early, and for that initiative — and, one must add, the resistance ever since to adopting the later-agreed-upon global standard for the task — it is now paying a technological price.
That is, if you consider being stuck with only one OS/software combo, and being stuck with dated standards a bad thing. But not keeping your software up-to-date, and having an unnaturally homogenous digital ecosystem, are both likely to leave the country vulnerable. You’d think that was learned on January 25th, 2003, but I suppose not. Though the prediction is that that Koreans will have insanely fast broadband (1Gbps) by 2012, one cannot help but wonder whether they’ll still be using Internet Explorer 6.0 as their main window onto the world… as well as how exportable those “new companies” will be, given their necessary focus on obsolete software and platforms.
(And though this may sound absurd, one finds it slightly more difficult to imagine that Korean IT community will suddenly wake up and recognize the problem: Internet Explorer 6.0 — a browser first released in 2001, close to a decade ago — actually got more users last month, bringing the total very close to 50% of Korean internet users. Apparently they used to blame Microsoft for the problems in ActiveX, going so far as to talk about suing the company in 2003 — even though Microsoft had indeed issued a patch for the vulnerability: one wonders who would get blamed if something like the Slammer virus hit Korea today. Maybe they’d do as the cops did in 2003, and blame China? Sorry, guys, but the Chinese didn’t force Korean ISPs not to patch their systems.)
But the ironic thing is, Korea was probably propelled into this situation by the advantage of a late start. It was easier for Korea to set up a wired broadband Internet infrastructure not just because of South Korea is small and densely-packed with people; there are a whole host of reasons, from the conscious promotion of the internet to housewives, the explosion of PC-Bangs (a “gateway drug” to the experience of using the Net, making it more desirable to do so at home), deregulation allowing competition to emerge, a nationwide impetus to change things up after the 1997 economic crisis, and even the desire to keep up with the Joneses. (More details on these and more causes are available in this report.)
Another reason, certainly, is that Korea didn’t have to go through the massive experimental period when everyone was faffing around, trying to figure out what was the best way to provide or receive internet service. There was competition, but the job of building a high-speed internet infrastructure on any real scale came years after the dust settled on the messy question of which types of hardware and wiring would work best (at least, for now… new cabling always becomes necessary eventually, but laying cable early on means replacing it if you make a wrong choice — which is easier since early choices are also more often relatively uninformed choices — always the curse of early adopters). By then, it was easy enough to choose a basic infrastructure model as the standard, and to even award construction companies for building structures (homes and offices alike) specifically to fit with the national standards for internet connectivity.
All this raises a couple of interesting questions, very pertinent to the next work of long fiction I plan to write:
- To the degree that a late start can provide an advantage in a given “complex technology,” (a technology with multiple tiers, where one tier can be cutting-edge while another is outdated) it can also create disadvantages further on down the line which proceed as a result of the late-start advantage. How is this applicable to technologies other than the Internet?
- Is it possible organizations — say, governments — to artificially retard the adoption of a technology or the development of an infrastructure until they can benefit from a late-start advantage? In “complex technologies” is it possible (or desirable) for a society to strategically retard elements of its development — certain tiers of a technology — to maximize late-starter advantages? (For example, had Korea waited the few months necessary to get 128-bit SSL encryption, would Korea’s internet be in less of an anachronistic bubble vis a vis the global internet?)
- How far does a lag need to go — or how long can it be ignored — before it is widely recognized that retooling is necessary? For example, Korean internet users are (a) primarily using Korean services and sites online, and (b) not particularly interested in the rest of the global internet. I can think of two scenarios which would prompt demands of a massive retooling: either peripheral services become impossible to access (say, a game that is released for a newer version of Windows only, and not for XP), or a catastrophic event (like a repeat of the 2003 Slammer virus incident) based on a vulnerability that strikes Korea especially hard (or, if we wait long enough, Korea alone, perhaps purposefully so) because of the country’s dependence on an obsolete OS. Which, of course, raises another question:
- Are there hidden benefits in using an obsolete technology similar to those of using an unpopular technology? (For example, the number of computer viruses for Mac, Linux, and other OSes are a miniscule joke compared to the viruses tailored for Windows systems. Will virus writers move with the world, working on new code to infest newer Windows OSes? Will any new viruses be coded for the obsolete OS?)
Interesting questions to play with. Feel free to comment, or not. I’m already thinking a lot of this through in terms of the novel I’m writing with the working titlle A Killing in Burma.
“Is it possible organizations — say, governments — to artificially retard the adoption of a technology or the development of an infrastructure until they can benefit from a late-start advantage”
Yes. Refer to HDTV and the US government in the late 1980s.
Good example! Been thinking some more about this, by the way.
Two things: first is that what was left out of Romein’s notion is that maybe the late start advantage might also have hidden costs — like, for example, the various social costs of Korea getting a full-fledged 1998-style Internet all of a sudden.)
I am also wondering about how it might work with some advanced form of hardware that, like software, is self-adjustable on the fly. Something more analogous to open-source software, using existing infrastructure and (relatively) self-adaptable hardware. Stuff like, whether a head start would or would not be an advantage then. Not sure…
If I had time, I can go on for pages. “Standards War” is a fairly important topic in international trade.
Korea right now has the best network available for processing customs clearance and trade related procedures. However, that may be in large part because the system is *new.* Countries which adapt later will probably incorporate latest changes so that in about 20 years, Korea will probably have an outdated network that it cannot fix without spending a lot of money. (Think of the classic example: QWERTY keyboards). This is not to say that Korea made a wrong decision making its network now. It’s just the nature of the game.
I also just want to mention that rather than smart “adaptive” connections, dumb connections may be better. Lawrence Lessig mentions that one of the reasons for Internet’s success is that the TCP/IP-based connection is so dumb tht new software have little problem adjusting to it. Parellel ports were fairly adaptive for its time, but since being replaced by USB, I can’t coonect any of my old parallel-port based hardware (one that really bugs me is ZIP disks. I have about a dozen zip disk saved with things I would like to keep – but it’s not enough to warrent buying a new USB-based ZIP drive. Are they even making those BTW?)
In the words of that immmortal philosopher and engineer, Captain Montgomery Scott… “The more complicated the plumbing, the easier it is to stop up the drain…”
Junsok,
This is not to say that Korea made a wrong decision making its network now. It’s just the nature of the game.
Yeah, this is the conclusion I’m drawing: there are all kinds of cost/benefit balances and it’s hard to say that a current benefit may not become a long-term cost, or vice-versa.
On Lessig’s point — hmmm. I am not sure whether it can apply to the technology I’m thinking about, which is quasi-universal fabricators. (Fabbers, but not like Trek ones… they can make things to template — products, like, say, a car body from compacted bamboo pulp or grass fiber, but not a banana.)
I think they need a minimum of “smartness” to function quasi-universally (building anything to spec, including copies of themselves), but the dumbness is that they are empty of models till you feed them. Throw in an Open Source archive of fabber templates and enough time, and you’d think economics would change utterly, right? (Well, sort of. I suppose trash salvage would be the new raw materials business.)
Anyway, the dumbness would be in the raw materials (not quite anything will do, but there’s a range) and of course the templates come via TCP/IP as well. But now I need to think about how the materials processing system would work. Hmmm. Got a few fuzzy ideas, but… maybe a kind of smart-dumbess would work there too…
Zip drives appear to be dead media, though I think you can get ’em second-hand (including USB-connective versions) from retro-computing dealers, or on ebay.
And, ha, Scotty. You Trekkie. ;)
Two things..
1) Since Korea can turn on a dime, when it finally wises up, I expect a quick move away from the silly shite to which they are currently wedded.
2) I detect an homage to Orwell in the title of your new novel. But I drink too much. ;-)
Charles,
To your first point, I don’t think turning on a dime is something Korea can do with everything. This change will take massive retooling, especially when it comes to standards of computer science education and standards of computer programmer accreditation.
We’re not just talking about a chunk of the internet, after all, but about a community of computer software professionals who have, for their whole professional lives, acted as if nothing but Windows existed. Show them Linux, and they ask, “Why aren’t you using Windows? Use Windows XP!” Show them Mac and their eyes cross.
We’re talking about a legion of people whose approach t software and to web design have been hobbled, suddenly adopting a system which requires much more work (and retraining, don’t forget), much more thought, and much more testing… the last, especially, being at odds with a professional culture where things are done at the last minute almost as a rule.
So I don’t think Korea really can turn on a dime in this case, more’s the pity. I really wish it could. I wish we could wake up one day and the announcement, “We’re going to unbreak our internet” could be made. But it’s not really likely.
(Besides which, most people I know here don’t really seem to notice that the K-net is actually broken. It’s fast, it has tons of animated JPGs, and they haven’t got much experience ordering from websites that don’t require you to sign up with your national ID, use a special “secret” card to access your bank account, and don’t use any other platform (so they don’t know how closed it is). Sure, some people know, but they’re mostly the people with a vested interest in preventing the change.
I do wonder, though, whether gaming will push Korea forward. Once the newest, coolest games don’t work on Windows XP any longer — once they require the newest Windows — then I can see an incentive build. Even so, one wonders whether Korea won’t insist that Microsoft retain backwards compatibility with ActiveX. It’s so ingrained now, it’s hard to imagine the whole meta-industry shifting without a nasty mess somewhere, after all.
As for the other point… well, sort of, though the two works you think I’m name-checking are actually in my to-read list. I actually came up with the title misremembering the title of a book by Marquez, as well as punning on the meaning of “a killing.” (Though, indeed, the killing does involve an elephant… of a metaphorical sort.) Also, I suspect there’s a song by The Cure in there, and something else I had on my mind just now but have lost… hmm.
I was told by a Korean acquaintance (a database admin) that one reason for Korea’s dependence on WinXP was Microsoft itself. According to him, around the beginning of Korea’s Internet expansion Microsoft set up hakwons across Korea to teach MS technologies. Graduates went on to fill lots of IT positions.
This is fine, really, since it’s hardly necessary to have a CS degree to do a lot of day-to-day Web/IT stuff. However, it would also mean that a lot of Korean IT professionals are relatively limited in their training.
Anyway, ActiveX applications are fast and easy to develop. Why make the leap to another method (Java, probably) when the established one works well enough and is fast and easy?
Rhesus,
Ah, well, now that is interesting. Of course, it begs the question: did Microsoft shut down its hakwons sometime prior to the time when it stopped promoting ActiveX?
(And you’re right, of course, about hardly needing a Computer Science degree to do “day-to-day Web/IT stuff” but the limitation in training is also a limitation in standards, in understanding, and in practice. People don’t even check what their websites look like on other platforms, or even in other browsers on the same platform. This is one of the most basic elements of “professional” web design, and in Korea it’s utterly absent.
As for ActiveX, there are plenty of reasons to stop using it. (One article out of many I could have linked.) Of course, those points only make sense to someone who knows it doesn’t have to be this way. In my experience, as with other issues, Koreans who don’t know this tend to assume it works the same way in other countries as in Korea. (ie. for example the signups with national ID, the endless downloading of ActiveX controls, the general instability of any website involving a financial transaction, etc.)
I was gonna say… ActiveX may be easy, but it is going to be gone and Korea is really the last country hanging on..
also.. I still believe Korea will be able to change quickly for two reasons –
1) When the government mandates something – boom – it gets done. Ask the cabbies about how their rigs are fueled.
2)Hours of work available. Koreans spend more time working (though not always efficiently) and the model is get it done (well, 85% done) ASAP.
I think when it comes, it will come quickly..
Sooner the better, really, since between ActiveX and Flash, Korean websites are a horrorshow..
Charles,
Yeah… However easy ActiveX is, Korea’s not going to be able to stay using Windows XP forever. Whether it’s global services, or gaming, or whatever, there will eventually be a pressure for adoption of a more global standard.
As for boom and getting it done ASAP: those are, to me, scary scenarios. Rebuilding your financial transactions systems from scratch… actually, has any country out there done it? I’d be curious how it’s done.
Also, I am pretty sure when it does get done, there’s still going to be all kinds of weirdness and horrorshowishness about the Korean internet, since web design is as much a cultural activity as it is a technical one.
Koreans are used to all kinds of flash and animations and audio that just comes on immediately and loops and so on. I don’t think they’re going to stop building websites that horrify us (or most of the Internet) even if global computing standards actually do get adopted.
(And I imagine we’ll likely still have major interoperability problems. Whatever gets designed will still only work with Windows, or only be tested with Windows, and so on.)
But it’d be nice if the move happened soon. Well, nice, unless it leaves the system vulnerable. One hopes they’ll hire consultants from outside to bring things up to speed, but…