Monday, November 17, 2008

Google's Banned Words

WHAT GOOGLE SAYS ABOUT SAFE SEARCH

"Many Google users prefer not to have adult sites included in their search results. Google's SafeSearch screens for sites that contain this type of information and eliminates them from search results. While no filter is 100% accurate, Google's filter uses advanced proprietary technology that checks keywords and phrases, URLs and Open Directory categories."


WORDS THAT GOOGLE SAYS ARE NEVER SAFE
Of course there are limit to even an "advance proprietary technology that checks keywords and phrases, URLs, and Open Directory catgories", so there some are words that Google thinks are never safe*, regardless of what site is using them, what site links to them, or what the context is. It doesn't matter if you're Wikipedia.org or a porn-spam blog, these words are never okay. Some of these words are not surprising. Some of them are surprising:


*Never safe means that if you set your Google SafeSearch to the "strict filtering" result, you will get no returns in a search for the string. What that means in the real world is that even though there are nearly 5,000 pages with the word "nude" on the Metropolitan Museum of Art Website, Google's "advanced proprietary technology" can't identify a single one of these pages as a "safe" return for the search [nude].

This is not an exhaustive list. It's just what I could think of and test. Maybe there are more.


WHO RESEARCHED THIS BLOG?

My name is Tony Comstock. I am an independent filmmaker. For the last decade or so, my wife Peggy and I have been producing and distributing an erotic documentary series called "Real People, Real Life, Real Sex." Each film is an erotic portrait of one real-life couple; how they met, how they fell in love, and what keep them in love.  Not surprisingly, sex plays a pretty important part in that equation.

Our films have played at film festivals and won awards in Australia, Canada, The Netherlands, Israel, the US, and maybe a place or two I can't remember right now. People like these films. They are warm and sentimental in a way that people don't expect coming from films that feature explicit sex. That's right. These films go "all the way" -- just the way that you go "all the way" when you're with your wife/husband/BF/GF/SO.

But because these film are unabashedly erotic and completely frank in their depiction of what it looks like when people who love each other make love we run into problems.

Stores in places like Australia, the UK, Utah, and Oklahoma that want to sell our films can't, because they're afraid they'll get arrested. 

Governments in places like Australia and Italy have threatened film festivals directors with arrest or having their permits revoked or their government funding cut if they play our films.

And then there's Google. I'm sure you've heard of them, right? They're the Silicone Valley start-up that's grown into the international conglomerate that more or less controls how people find things on the internet. They way they do that is through some complex algorithm that looks at the words on your site, and the people who link to you and what they say about you, and then tracks that back to what people type into Google's search box.

If you ask me, that's pretty fucking clever, and the rest of the world seems to think so too. Google's market cap is somewhere north of $90B, making Google about as big as the world's biggest auto company. Wow!

The problem is – not for Google, they're doing fine; for me and my films – is the way Google's clever algorithm works, it end up giving more "algorithmic deference" (I think I just coined a new phrase!) to places like stormfront.org or the americannaziparty.com than it does to our little corner of the internet.

Some of this is simply a product of popularity, which I can accept. More people write about and link to stormfront.org than write about and link to comstockfilms.com. Thats' just the way it is.

But some of it isn't. Some of it is a product of the language that we have to talk about sex, or more accurately the limits of that language, and how Google's algorithm treats words that relate to sexuality. In Google's algorithmic world, some words are "safe" sometimes in some places, depending on who links to those place. But some words are never "safe." 

What does Google means by "safe"? "Safe" for whom? "Safe" from what? Who can know? But knowing what the "never safe" words are is a start. If you find a "never-safe" word, please let me know.  E-mail tony at comstockfilms dot com, subject never safe word.

Thanks!