When autocorrect goes horribly right

There were a few things that were sweetly grandmotherly about Patricia Duncan Moran's Facebook post to her great-granddaughter, Devyn, on her birthday. For starters, Moran had posted it to her own wall instead of Devyn's. She named but did not tag her great-granddaughter. But most striking was the signer: "Happy birthday to my beautiful great-granddaughter," the message read. "See you soon. Love, Great Grandmaster Flash."

Great Grandmaster Flash?!

It was the PG version of what's come to be known as the autofail, the accidental autocorrection from which many a blog and book have spawned. In this case, Facebook's autocomplete feature had made the assumption that "Grandma" was short for "Grandmaster" and adjusted accordingly, tagging the hip-hop pioneer in the process. "It became quite the family joke," Moran, 83, says on phone from Montreal. "I kept saying, 'Who is that?' because I had never heard of Grandmaster Flash. But now everybody calls me Great Grandma Flash."

Botched autocorrects are a byproduct of a technological convenience that allows typing on the go, even when the message does not always come out as planned. Yet, as autocorrect technology has become more advanced, so have its errors.

It's not simply that a town called "Cupertino" (the home of Apple) may appear when you are trying to encourage collegiality, as was an early default of Microsoft Word. Nor is it that similar letters may yield an unwanted "prostitute" sandwich. Nope, today's autocorrects feel almost personal.

There was the friend, a cocktail lover, who texted her pediatrician to inquire if she should switch her 2-year-old from "1 percent milk to 'whiskey." (He said yes, definitely, she should.) And Allyson Downey, a New York entrepreneur whose frequent response to something she liked - "Love 'em!" - always seemed to read "Love me."

And then there was Naomi Campbell, who sent a tweet to 300,000 followers congratulating "Malaria" (that is, Malala) on winning the Nobel Peace Prize.

Autocorrect originated with word processing programs of the 1980s, in which the language used was checked against a dictionary to make sure the spelling was correct. Back then, the point was simple: to make typing faster and more accurate. To help you, you know, not look like an idiot.

But these days autocorrect is creating problems as it solves others. Tech companies like Google, Facebook and Apple employ dozens of linguists - or "natural language programmers," as they are known - to analyse language patterns and to track slang, even pop culture. And they can do amazing things: correct when you hit the wrong keys (the "fat finger" phenomenon) and analyse whom you are texting, how you have spoken with that person in the past, even what you've talked about.

Apple's iOS 8 operating system, released in September, even purports to know how your tone changes by medium - that is, "the casual style" you may use in texting versus "the more formal language" you are likely to use in email, as the company put it in a statement. It adjusts for whom you are communicating with, knowing that your choice of words with a buddy is probably more laid-back than it would be with your boss.

Your smartphone may now be able to suggest not just words but entire phrases. And the more you use it, the more it remembers, paying attention to repeated words, the structure of your sentences and tone. All of which is fine, except that it turns the notion of the guiltless autocorrect on its head. These days, autocorrections are likely to tell the person on the receiving end something about you. More recently, Google was criticised when a web search for "English major who taught herself calculus" suggested changing "herself" to "himself."

Johan Schalkwyk, an engineer who leads speech efforts at Google, said, "Keeping up with slang and trending acronyms is like a jungle" - a jungle full of cultural land mines.

But of course, the more we fail, the more understanding we have about those fails - or at least we hope so.

©2015 The New York Times