I have a friend who, years ago, used to go about inserting subtly absurd photos into Wikipedia articles.<p>One that I recall was for the Refrigerator entry, which he seemingly innocently updated with a "better quality photo" of his own fridge.<p>If you examined the photo you'd spot that there was a guitar on one of the shelves[1]<p>[1] <a href="http://en.wikipedia.org/w/index.php?title=Refrigerator&oldid=130709805" rel="nofollow">http://en.wikipedia.org/w/index.php?title=Refrigerator&oldid...</a>
The article referred to him being "name dropped into books". I wondered if the article meant "scraped into the awful spambooks that sell Wikipedia content without the copyright notices", but was amused to find out it actually made it into a list of abandoned deities compiled by a serious-sounding professor of philosophy for a book on atheism
<a href="http://internet.gawker.com/how-one-man-made-himself-into-an-aboriginal-god-with-wi-1692426415?utm_campaign=socialflow_gawker_facebook&utm_source=gawker_facebook&utm_medium=socialflow" rel="nofollow">http://internet.gawker.com/how-one-man-made-himself-into-an-...</a><p>Considering the article was an unsourced stub that seems to have owed its longevity to the fact nobody noticed its existence, that's some impressively <i>comprehensive</i> sloppy research.
I find it a little weird to see Gregory Kohs listed generically as "a prominent Wikipedia critic". In 2007 he tried to start a business where he'd get paid to put articles in Wikipedia:<p><a href="http://www.nbcnews.com/id/16793247/ns/technology_and_science-tech_and_gadgets/t/idea-paid-entries-roils-wikipedia/" rel="nofollow">http://www.nbcnews.com/id/16793247/ns/technology_and_science...</a><p>At the time it was seen as an outrage (in my view correctly so) and promptly banned. Wikipedia has a clear conflict-of-interest policy [1] for good reason. Their most valuable asset is reader trust, and Wikipedia, for all its flaws, has done a reasonably good job of keeping it from being overrun by marketroids.<p>[1] <a href="https://en.wikipedia.org/wiki/Wikipedia:Conflict_of_interest" rel="nofollow">https://en.wikipedia.org/wiki/Wikipedia:Conflict_of_interest</a>
Australia is made up of hundreds of indigenous lands and language groups many of which carry sacred knowledge not shared with the broader public. So it's not totally surprising that this hoax survived as long as it did.<p>To give you a fascinating extent to how varied indigenous lannguage groups of Australia are check out this map. We're not talking dialects/accents here, we're talking unique languages where I've be told in many cases people share very little traditionally language except for possibly some similarities with their direct neighbouring lands.<p><a href="http://www.abc.net.au/indigenous/map/" rel="nofollow">http://www.abc.net.au/indigenous/map/</a>
The situation will only get worse as the number of editors declines, leading to a death spiral where hoaxes, bias and inaccuracy make readers and editors lose faith in the project. The other problem (as noted below) is that it is almost impossible to add new or updated information. Generally, observers blame the community, but, in fact most of the site's problems are symptoms of poor software design. I write more about these problems here: <a href="http://newslines.org/blog/wikipedias-13-deadly-sins/" rel="nofollow">http://newslines.org/blog/wikipedias-13-deadly-sins/</a>
This is both funny and extremely frightening.
Between this and some links from other HN users, I feel like wikipedia is rapidly losing credibility.
Does anyone have recommendations for any alternatives ?<p>Personal Anecdote:
I also remember the time, one of my friends in college told me that she purposely added misinformation to Wikipedia the night before a big test to lower the curve.<p>That was also the moment I was glad I'm didn't do premed in college, those kids were hyper-competitive and unethical.
Machine learning will fix this. Currently it's likely the case that more info is entered / updated than editors can handle, but once machine learning systems point them toward "suspect" entries they'll be able to keep up with everything I imagine.