Stop Abusing AutoCorrect
“There is a right and wrong way to use your phone’s autocorrect,” says the inventor. Ken Cocienda, former chief iPhone software engineer for Apple and author of the new book Creative Choice: Apple’s Design Process in Steve Jobs’s Golden Age , led the team that created the first soft keyboard for the first iOS release. And he says that with auto-correction, you can get too aggressive.
AutoCorrect has evolved since its inception, but it is largely based on the same fundamental factors. When you enter a word, your keyboard software asks the following three questions:
- What vocabulary words, if typed “correctly”, will look like click patterns for the user? (Your keystrokes form a “shortcut key” that your phone compares to the ideal shortcut for each dictionary word.)
- Which of these words are more common and how much ? (Cocienda points to Zipf’s Law , which predicts the proportional frequency of words in any language.)
- Which of these words appear in the user’s “dynamic dictionary” – their most frequently used words, or their contact names, or words they retyped even after autocorrect changed them?
For each word, the program must choose how much to weigh these three factors in order to decide what you mean. To understand how difficult this choice is and how much is at stake, remember how long it took to teach your phone the word “fuck.”
But this whole process works best if you don’t stop correcting every word. It is much more efficient to type a whole sentence or message and then subtract it at the end. This is what the system was designed for and what works best in Apple’s user testing.
“And this will become even more true as autocorrect and predictive text become more complex,” says Kocienda. As the keyboard explores longer passages, it can recognize more errors. For example, he says, your phone might think you typed in, but implement a few words later on what you meant when you type. (He knows you didn’t mean, “You love Pina Colada and get caught in the rain …”). The more advanced the auto-correction function, the more it can fix the situation, even if you don’t watch it constantly. So learn to trust autocorrect. (Or turn it off .)
And don’t rely on predictive text all the time. I use this bad habit too often by checking the suggested words above the keyboard to see if I can hit the finished word to save a couple of seconds here and there. Cocienda calls this context switching: you type, then read, then watch, and every time you switch between these tasks, you waste time and attention. In Apple tests, most users (not just experienced typists) typed much faster if they just completed their words. Smart text is best ignored except in unusual words or situations.
This revelation reminds me how much I miss the BlackBerry and Sidekick hardware keyboards. But, as Kocienda points out, since soft keyboards can disappear and make room for dual screens, they are essential for full screen phones. In his opinion, we could not get the full-featured applications that we have while lugging around with the hardware keyboard. Therefore, we have to give up the simplicity of pressing a real physical button, which in fact was not so easy.
Kocienda tells Lifehacker that what sets Apple apart from competitors like Google is that Apple always strives to make the single best choice for most users, rather than passing the choice to users, which can overwhelm busy people who just want to send text. This means that some features and options appear later on the iPhone than on Android, but as soon as they appear, they are usually thought out to the smallest detail. ( Usually. )
Kocienda says this is another reason technology needs diversity: devices like the iPhone are for everyone, and you can’t design for everyone unless everyone is represented on your design and engineering teams. You cannot design a software keyboard without considering the size of the hands and fingers that use it. And if you don’t expect problems until you start user testing, you won’t be able to find better solutions.