Are we running out of tweets?
With 26 English letters, ten digits, and 33 special symbols, we have on the order of 70 unique characters to tweet with (ignoring upper and lower case and other languages, of course). The total number of possible tweets can thus be calculated by raising 70 to the 280th power. This is a really big number. How big? For the non-exponentially inclined, that’s basically 4.24 followed by 516 zeros.
But this estimate drastically overstates the number of tweets because the vast majority of those random combinations of letters would never be typed by any human being, with the possible exception of Donald Trump. So how do we limit our count to only tweets involving legitimate English words?
There are roughly 170,000 commonly used words in the English language. Let’s exclude the words that only Ken Jennings knows and reduce that to 20,000, which is thought to be, roughly, the size of the average English speaker’s working vocabulary.
The average length of an English word is about five letters and let’s ignore the fact that young people like to use short words that people like me have to look up on Urban Dictionary (“u r my BFF, LOL”). Next, let’s divide the available 280 characters by six (five letters plus one space character after each word), which gives us ~47. Thus, we could approximate the number of tweets using only legitimate English words by choosing up to 47 words from a pool of roughly 20,000, which is 20,000 to the 47th power.
