It's no secret that the Internet can sometimes be a pretty rough place. Some people enjoy making others feel awful (see How Trolls Work for more on that). But it's also surprisingly easy to discourage other people, even if that's not your intention. That may be one of the contributing factors to the decline in the number of editors over at Wikipedia, and artificial intelligence might be the solution.
The underlying problem is the same thing that makes Wikipedia such a power tool — it relies upon a community of editors who write, shape and edit all the articles contained in the online encyclopedia. But editing an article on Wikipedia isn't easy. The default editor uses a complicated markup language, and there are many rules that govern how editors can make changes. If you edit an article and break a rule, chances are someone will come along and reverse your edit without notifying you.
To new editors, this can be disheartening. If the edits you make are consistently struck down, you may conclude that you're wasting your time. Worse, you might feel you're being treated like someone who is intentionally vandalizing Wikipedia entries. For Wikipedia to survive, this can't continue. According to MIT's Technology Review, the site has lost 40 percent of its editorial community over the last eight years.
That's where the AI comes in. It's called the Objective Revision Evaluation Service, or ORES. It can flag editorial changes that don't comply with Wikipedia rules. Beyond that, it can determine if the person who broke the rules did so as part of an innocent mistake, or if it was the act of a troll making “damaging edits.”
Reviewers will be able to send a message to the editor in question, alerting him or her to the problem. In other words, the AI's job is to remind human editors to show a little humanity toward others. Perhaps this will prevent new editors from feeling dismissed by the old guard and help Wikipedia build up its editorial community again.
It's not mandatory that the Wikipedia community use ORES when reviewing articles. The software won't magically soothe the hurt feelings of editors whose work has been deleted. But for Wikipedia to survive, the community will have to find ways to encourage new members and retain their numbers. Otherwise, the crowdsourced encyclopedia will encounter problems with errors, outdated information and obsolescence.