Tuesday, February 01, 2005

Devaluing News with Google

Ever since its launch, a mythology has grown up around Google News based on the way it's been hyped by people who ought to know better. The hype covers a couple of misconceptions which are driving a fairly complete mythology about Google News, most recently in my eyes seen in a short video published here:
oak.psych.gatech.edu/~epic/

The misconceptions undermine not just the popular understanding of news, but help create a world in which information is dramatically devalued.


I'm not going to try and trace the history of the myths; instead, merely for simplicity, I'm going to draw them completely from the "Epic" video and keep things moving along.


Myth: "Google News is edited entirely by computer".

Reality: "Entirely by computer" is a misconception. The choice of sources trawled by Google is made by people, and they're starting to suffer under the overload of competing sources. They are trying to create rules to assign "validity" to source, some of which are stupid and arbitrary, and these "validity" decisions are not made "entirely by computer".

As a result, the observant can already see a deterioration in the quality of Google News, at least in three directions:


* syndication is a kind of "keyword spam" in which a press release, picked up automatically by Website fillers, becomes the dominant angle on a story. This is a serious problem, because the author of the press release - almost never an objective source - can skew the Google News view of the world;


* the savvy or cynical can already "game" the Google News system, I don't know how; but when The Inquirer (for example) can't get Google News listing but FAQs for games are treated as news, there's a problem; and


* the Google News decisions, such as they are, are arbitrary and opaque. One news source I know was delisted by Google News merely because it relocated its host; there is no avenue for approach or appeal, no editor, and no access to the decision-maker.


The other reality associated with the above myth is that Google News is not edited at all. It's not edited by people, it's not edited by computer; and to imply any kind of editorial process is to assign to Google News a credibility and reliability which it doesn't have.


Myth: News can be "constructed entirely by computer"
This is already a growing myth among futurists, and with the huge amount of stuff associated with the Google index, is seems logical that the right ruleset would let Google do the cutting and pasting, right?

Wrong.


First, "entirely by computer" ignores the human authorship of the source material. Even if a bot can assemble plausible content by grabbing sentences from somewhere else, the ultimate author is someone.


Second, the myth ignores the only way in which a reporter can add value: by going beyond the press release.


In the Epic view of the world, news is constructed out of press releases. The problem is, Google News is fostering that very belief. The reason for the "syndication spam" I mentioned earlier is that in the online era, news sites seek hits-above-all-else; and this, regrettably, means the editors - the real humans, I mean - have taken a group-think decision to get syndicated wire pieces online as fast as possible.


And this is the real hell, and the real output of Google News: that the information people receive is, far more than even four years ago, dominated by the press release and the syndication.

Even when a journalist produces an outstanding work, the Google News impact is to devalue the original.


Instead, a top-notch story will get condensed by a wire, redistributed as "according to a New York Times report", and it's that syndication rather than the original which will dominate the Google News search.


Real news gets exactly the same weight as a press release.


What Google News offers in the long term - the seeds of which you can already see - is a kind of placebo news built on press releases.

As to the kind of academic visionary who equates the media release with the news, what can I say? Most academics form their opinion of journalists based on snobbery: their training is better than the journalist's. Hence the journo has no real value - since we merely reproduce press releases, why not get a machine to do it?

It's a pity that such a view gets currency in the real world...

No comments: