When I created my microblog post, I expected it to be complete nonsense. I was surprised to see that it actually started out sounding like a real post! Then I realized exactly why this was happening. You see, my school division is one of several divisions in the province who had to switch to a new student information system this year. As a part-time learning consultant, my main job this past couple of weeks was to help teachers and administrators with this new program as our first term just ended on Friday.  My cellphone (which is the device I used to write this message) also doubles as my work phone. This means that over the last few weeks I’ve sent countless messages containing keywords like: time, work, easier, better, and kids. It’s no wonder these words appear in my predictive text!

The only places I’ve really noticed a predictive text message being written has been in microblogs or text messages. I don’t think people use predictive text regularly and it’s hard to tell if the posts I’m thinking of were just accidentally changed by auto-correct, or if they were created with predictive text. I noticed that my predictive text message does not sound like me. Sure it contains some keywords I’ve been using, but my writing tends to be a lot more expressive than the message you see above.

Even though algorithms can pick up on frequently used words, and generate the small words like: is, a, or and after specific words, I still don’t think predictive text is something that will be used frequently. As a teacher, I’ve had students who struggle with writing use a predictive-text feature on their computers because it allowed them to just select the word they wanted rather than trying to spell it out. However, with recent improvements to the voice-to-text features, these students don’t even use predictive text frequently anymore. While algorithms make all kinds of artificial intelligence possible, I don’t see them becoming more than an intelligent spelling and grammar checker in any large piece of writing.