It's not funny, Google — a Philosophical Analysis of Search Engine Algorithms and Social Media

In a recent article on Mashable, they talk about the newest algorithm of Google that supposedly ranks the funniest videos on YouTube. "Can humor be quantified?" they asked. Righteously so, I've been very doubtful of this for a long time just as when I compared the search engines to IBM's Watson.

In high school, our philosophy professor discussed John Searle's Chinese Room Argument that lays out the differences between syntax and semantics. The first being the ability to function based on preprogramed or predefined impulses and catalysts and the latter dealing more with Qualia or the quality of understanding or comprehending subjective experiences.

So what is it that makes something funny? Surely it is subjective: what is funny for me might not be funny for you. But the real question would be can algorithms unlock this quality of funniness? Are there patterns in human behavior that search engines can be programmed to look for?

First of all, as I've mentioned countless times, the search engines can't even really tell what is in a photo or a video. It bases everything on keywords on the tags and/or content that surrounds the mentioned media. For example, if I upload a random video of a bouncing ball and put a title of "funny guy" and stuff it with tags synonymous to that, the algorithm has no choice but to think the bouncing ball is a funny guy. Yes, you can beat the search engine.

Rebeca Black went viral through social media, not algorithms

And so they put in the human factor: the democratic process of voting whether you like or dislike a video. They knew that people's love or hatred for a video is a big indicator of the video's quality.  They also consider view count, number of comments, and number of times it has been shared through social media. Although it has significantly improved search results, popularity does not necessarily equate to quality. 

Let's take Rebecca Black, for example. How would you quantify the quality of what makes her song Friday entertaining? Can algorithms really determine that quality? It's just so unknowable. Rebecca Black's Friday will forever be the mystery that bewilders our existence.

In the Google Research Blog, they discussed that they added how many times people say "LOL" and how they say it to the funniness algorithm. Haha! That's almost ridiculous. They seem to be running out of options. It is inevitable for them to return to the concept of what has made Web 2.0 great and  that is, as I have mentioned, democracy—giving the power back to the people. This is exactly what they want to achieve through their planned "Search, plus Your World" campaign:

Is Google giving up?

I bet Google hates it that they are not the ones leading in the social media industry. Google is hungry for Facebook's database. I've made a strong case for this when I wrote about Why Facebook has an edge over Google in Advertising. They are working so hard in waiting for Facebook's tipping point and planning on how they can take over. As you can see in the video above, social media, letting the people decide, is the best way they can improve search. They are certainly very far from that though.

Of course, I do not reject the idea that one day, the algorithms can actually be so powerful and knowledgable. Maybe one day we can ask Google the meaning of life. Maybe the search engines can tell us what is love or beauty. For now, such mysteries of life are left to our own discretion. Knock yourself out.

No comments

Just another one of the blogs of Harry Santos. Powered by Blogger.