Wikipedia has come a long way from its early days as an open source encyclopedia that your high school English teacher taught you never to cite from. As the site has grown more affluent and dedicated to fact-checking, some…curious side effects have been observed.
Researchers at the Oxford Internet Institute sifted through nine years of data about automated editing programs at Wikipedia and found some amusing patterns. Specifically, many of the bots have been in personal editing wars, often against other individual bots. Publishing their findings Thursday, it seems amusingly that machines have gone to flame war, editing and counter-editing the same minutia for years on end before anyone caught it.
Wikipedia uses many automated programs, commonly “bots”, to parse through the site’s millions of pages and make small corrections to bring a page’s wording into line with the website’s standard, or to catch obvious attempts to sabotage a page. But most of these bots are user created, so once approved by Wikipedia, there is little linking them to each other.
The end result is the various bots going in to make minor corrections on web pages, only for another bot to scan the same page, detect an “error” from the previous robot, and correct it back. Over time these conflicting scripts give the impression that specific bots are targeting others, almost like a form of internet harassment. These findings could give us some insights into the internet as a whole, as on major sites like Twitter, bots generate almost 25% of all tweets.
While the machines may not rise yet, we should keep a wary eye on all the practice they are getting. Even if it is only trolling each other.