Monday, March 2, 2015

Everything on the Internet Doesn’t Have to be True: an age-old argument influencing a new idea


DISCLAIMER: The statement that makes up the title of the article, “Google has developed a technology to tell whether ‘facts’ on the Internet are true”, is ironically (and hilariously) false.  I’ll explain in the post.


Not everything on the internet is true.  This is one of the first things children are told when they are introduced to the internet.  People post misinformation on the internet all the time, whether they don’t know what they’re talking about but want to sound smart, or they just want to misinform people just because they think it’ll be funny.  Sometimes, people online post misinformation just to get attention; this is especially true with many news outlets, whose writers will lie in the very titles of their articles just to get more people to read what they have to say.  There’s an immediate example the form of the article I chose to base this post on: the title of the article is “Google has developed a technology to tell whether ‘facts’ on the Internet are true”, but in the body of the article, they immediately acknowledge that it’s “100 percent theoretical: It’s a research paper, not a product announcement or anything equally exciting”.  With that example of my point aside, however, the article (and the research paper that it mentions) brings up an interesting prospect.

In the research paper, a team of computer scientists at Google proposed a new method of ranking search results.  This new method would be based on the factual accuracy of the web page, as opposed to its popularity, and its inner workings are detailed in the article.  This is an interesting prospect: to be able to distinguish websites containing accurate factual information from those filled with lies would definitely help people come up with more accurate and reliable Google searches.  This would be a boon for people who use Google for research, and for anyone who wants factual information.


But it wouldn’t be without flaws, and at any rate, it’s probably a long way off.  This sort of filtering and identification process would require a massive database for cross-referencing information.  And then, of course, there’s the matter of minute details: small, barely noticeable points hidden in blocks of text.  Barely noticeable, that is, unless someone’s actually reading the page in depth.  Depending on how much detail there is on the webpage, there can be so many words, so many ways those words can be used, and so many things that they can mean depending on context, I just don’t believe it’s possible that every detail on every page can be verified.  That probably isn’t entirely necessary; Google’s theoretical truth-finder should be cross-referencing keywords and headline words, and should help show people the most truthful results based on those things.  But the point is, if such a system is put in place, there’s no way it’d be completely foolproof.

No comments:

Post a Comment