As many in the industry have been doing, I've been exercising Google's SafeSearch feature in concert with Google Instant to see what's suggested and what's not (many bloggers dream of finding some terrible results - always makes for a good linkbait article). I have to say, for most words one would expect, Google Instant is doing a pretty decent job. However, I tried searching on "google is e" (thinking "evil" results would come up) and got no suggestion. Not only were there no suggested alternatives, the entire search result page went blank:  I then tried every letter of the alphabet - they all returned instant results. Then I tried "george bush is e", "britney spears is e", etc. - no instant results. For whatever reason, Google Instant seems to have miraculously eradicated "evil" from the world (!) This seems like a slippery slope on Google's part, if some sort of override was put in driven by possible libel concerns, eradicating such results seem like an impossible task. Or are they perhaps using some sort of automated sentiment analysis? It doesn't seem so; various people for example are still apparently "an idiot" (for [...is an i]), and surprisingly some are "satan" (for [...is s]) as well as many other negative phrases. Has anyone noticed any other non-SafeSearch oriented self-censoring in Google Instant? Please limit any examples to words which are clearly not obvious SafeSearch examples, and keep it clean - thanks.