Dave has an interesting post today about the problem with centralizing services on the web. He goes on to analogize this to the nature of “news” on the web. I think his first point is valid. It’s a major premise in much of computing that a single point of failure can cause an awful lot of problems. Whether that failure is technical, personal, or political makes absolutely no difference.
Arguing against centralizing users and data in one web service is easy to defend. Media is different.
Dave criticizes TechMeme because it’s the quintessential centralizer. What Dave neglects, though, are the benefits of centralization in the media.
In a web-service, the benefits of centralization are few: uniformity, economies of scale, a larger knowledge base, etc. These are trivial benefits that are far outweighed by the likelihood of a single point of failure, in fact, failing.
The major flaw of the centralization of media isn’t nearly so bad. The argument is that, when media starts to congeal, important things are left out. The de-centralized nature of publishing platforms, ironically, is what so greatly mitigates this problem.
Also competing to make these two examples so different are the BENEFITS of centralizing in the media. This is what people do. We categorize, we group, we drill-down. We’re notoriously bad at, for example, reading 500 blogs a day and trying to discern the interesting from the non-interesting in an efficient and effective way.
Places like TechMeme are for news consumers. They’re for people who want the interesting stuff distilled out for them. If you’re not one of those people, and I don’t think Dave is, TechMeme isn’t going to be terribly useful to you.
Dave alleges that bloggers have started to write for TechMeme, to try to draw readers (presumably to try to increase their ad revenue). If we assume that this is true, it could easily be seen as a problem. When the human filters stop trying to filter and start trying to bandwagon, things WILL get left out.
This is where the argument comes full circle, though. The problem is not the filtering of the headlines… by the TechMeme’s of the world, the problem is that TechMeme is the “single point of failure” in this news-filtering web service.
More TechMeme-like services, if they existed, would need to differentiate, so they would have an incentive to find and highlight things that are DIFFERENT from those at TechMeme. Give those services some market share, and now the incentive to cater to TechMeme when posting has been greatly reduced. In this hypothetical world, you just do your own thing and hope one of the TechMeme-type services finds you… but you know that trying to cater to any one of them would all but guarantee that you won’t make it onto the others.
In the end, categorizing and filtering news is an invaluable service… but that service needs to be decentralized to mitigate or eliminate the perverse incentives. Interestingly, Dave got the problem right… but it’s a second-order problem, not a first-order problem.
UPDATE: The latest on the Loic Le Muir story arc… Dave is going to a conference is Paris that Loic is running (I think) and Dave is an INVESTOR in Loic’s start-up… as is Mike Arrington.
Mike disclosed the investment… and then Dave did.
So, Dave will protect his investment. I suspect he’s placed a few other bets recently… he’s very kind to Fred Wilson of late. That’s a bit of a clue.
There’s usually a benefit for Dave’s kindness… a benefit to Dave.