2. The heartbreaking hilarity of @GooglePoetics: http://www.googlepoetics.com/ 
4. A 2011 interview with @jonwiley on being a Google Autocompleter.

“On a good day, I average 34,000 words per minute. And I go through a new keyboard every eight days.”

5. History of the predictive text swearing:
(A skit apparently much more grounded in reality than you might assume: https://www.urbandictionary.com/define.php?term=chubi&defid=1907493)
6. Words censored on various forums because of uncouth words hiding within:

assassin
cockpit
canal
screwdriver
Saturday
hearse
basement
farther
skyscraper
grapefruit
manuscript
William Shatner
Bumblebee
Riddick
cluster
Lightwater Town
circumvent
suspicious
reputation
Japan
(Otherwise known as “S  cunt  orphe problem”: https://en.wikipedia.org/wiki/Scunthorpe_problem)
Isn’t this all getting too cute and meta.
7. Google’s “did you mean” campaign in 2010.

(I feel there must be a million jokes funnier than the ones they chose?)
8. Must be one of the earlier “Google’s algorithms reflect the sexist biases of its creators/society” articles, from 2009:

https://www.salon.com/2009/10/23/google_fail/

/cc @sara_ann_marie
9. Even more funny since I was the designer on this very feature at Google.
10. A few years ago, Mashable made animations for weird, unexpected Google suggestions.

http://mashable.com/2014/07/13/google-autocomplete-funny-illustrations/
11. Google Feud! A Family Feud-like game (for one) where you guess Google suggestions: http://www.googlefeud.com/ 

(It’s really hard, at least for me.)
12. “Google, democracy and the truth about internet search”

Just a year ago, a search for “are women” autocompleted to “are women evil?” and immediately showed a result saying “Every woman has some degree of prostitute in her.”

https://www.theguardian.com/technology/2016/dec/04/google-democracy-truth-internet-search-facebook
(This thread became such an emotional rollercoaster.)
14. And then there’s this: https://twitter.com/markedly/status/943188943977238528
15. “We used predictive keyboards trained on all seven books to ghostwrite this spellbinding new Harry Potter chapter.”

https://twitter.com/botnikstudios/status/940627812259696643
(Not too into Harry Potter? How about Arrested Development?) https://twitter.com/botnikstudios/status/943542693401120772
17. Today’s word: cacography, a “deliberate comic misspelling, a type of humour similar to malapropism.” https://en.wikipedia.org/wiki/Cacography 

E.g. teh awesome, hueg like Xbox, and so on?
18. The very first, 1993 edition of AutoCorrect in Microsoft Word only did nine things:

(r) » ®
adn » and
don;t » don’t
i » I
incl » include
occurence » occurrence
recieve » receive
seperate » separate
teh » the
(By next time in Word 95, the AutoCorrect list was much more vast, including support for the most haggard-looking smileys in human history.)
(And, did you know that modern Word has extra autocomplete FOR NERDS? I didn’t!)
(Funnily enough, I just watched my brain autocompleting “autocorrect” to “autocomplete.”)
19. Speaking of Word 95, this is when the (in)famous red underline appeared for the first time. In marketing Microsoft momentarily branded it as “Spell It,” but on day one the UI already simply said “Click the right mouse button on the red wavy underlined words.”
20. Heh, but also sigh. https://twitter.com/nathanaelb/status/891089433667096576
22. Just wanted to share this for posterity… the first, original Word 6.0 AutoCorrect dialog box with the glorious *nine* entries, in 100% high fidelity.

(I only saw it in a scan of a book before, so I installed Windows and Word in an emulator to get this.)
23. We might never learn this lesson / This might never cease to be hilarious: https://twitter.com/badmanbegins/status/956630644838359041
24. :facepalm: https://twitter.com/itsnicolenguyen/status/1022995740287614977
25. https://twitter.com/crushspread/status/1054704019380162560?s=21
27. Deep inside my public library, I found this oldie, but goldie from 1994.
28. https://twitter.com/pegerella/status/1069382158211575809?s=21
29. Sort of the inverse of some of the above, but this entire thread is too good to pass up. https://twitter.com/cabel/status/1072976837783904256?s=21
(I’m beginning to see this thread as obligatory reading for anyone who ever tries to do smart things with people’s typed text.)
31. https://twitter.com/innesmck/status/1060294661171015681
33. Spotted while walking in San Francisco.
34. Spotted in San Francisco, too!
35. A fun video about the Scuntorphe problem from @tomscott:
37. https://twitter.com/jmspool/status/1263807353788542976
39. https://twitter.com/greenhousenyt/status/1271273745890639872
40. Somewhere at the top of this thread I said “fun,” and I am not living up to this promise. https://twitter.com/Guinz/status/1280698069932277760
You can follow @mwichary.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: