I just tried to use google translate in hebrew; giving it words in all lowercase provides a gramatically very different translation than giving it words all in uppercase. For example, typing "good morning" gives the hebrew equivalent of "it is good in the morning" (טוב בבוקר), while typing "GOOD MORNING" gives the appropriate greeting in hebrew (בוקר טוב). Not sure if that's a bug or a feature, but definitely very screwy.
I like "I HATE WATER"
http://www.translationparty.com/#8642238
"If you are not used to using water that is being used to sign for you, I do not sign up for Facebook."
Reading all the other answers with links to translationparty.com, I decided to google "translation party wikipedia" (without the quotes) and the first result was "Ranks and insignia of the Nazi Party - Wikipedia".
I happened to see meanwhile that even if I type 'goobe'(which means 'owl' in Kannada) after setting source to auto, it says "We are not yet able to translate from Kannada into English". Really, amazing!
I don't think this is an easter egg. Googles translations gives the user the ability to help them translate things better to future use. Thinks it's a bug. :)
Makes me think if machine translation is to improve it will almost certainly require human translators to contribute. The style in which Google is doing this seems like this will happen more often.
「私はあなたのFacebookを利用しています。」 is a grammatically correct but awkward way of saying "I am using your Facebook."
Edit: Seems they omitted the の and my brain added it for me. Whoops. That makes the sentence quite awkward indeed: the most natural reading is the one I gave, but you could read it as "I am using you, Facebook."]