
This is a matter likely to reopen the debate on Twitter: some see it as further proof that the content generated by users are not a reliable source. Others marvel at this time so short (2 min can be used even if a fan to a heart attack) and note that the "wisdom of crowds", definitely, is greater than "some intelligence experts."
Twitter: automatic corrections

It made me want to dig how Twitter works and how the system produces the highest reliability possible for its readers. I knew that you or I could contribute freely, but I did not know that since 2007 a robot watching the sweet name of Salebot . Salebot been in operation since October 21, 2007, and removes about 200 changes per day (6000 per month). Salebot has a learning function and has no equal to uncover (and correct immediately) whole sections of text disappearing characters or words to filthy or offensive. One can imagine it, Wikipedia tells us, the number of people who find an interest in drag "pee poo" in the middle of a reference text.