For this week’s task, I decided to use the website Tweetgen in order to make it easier for myself to deal with the 280 character limit. I don’t have a twitter account, so this “fake tweet” website was useful as it allowed me to see the character count as I typed. Check out the video below to see how I completed the task:
The final message came out as “As a society, we are going through the same thing that we can do with our life. I just got a new idea for my next step in my life.”, which ended up being 130 characters in total.
I found this activity both interesting and frustrating. It was interesting to see the types of words that the predictive text algorithm generated. It is fascinating to me how quickly the software does the necessary calculations and computational processes in real-time to try and predict what the user wants to say. The software is clearly trying to decipher my intentions using contextual clues as well as drawing on the billions of words and phrases that have been typed at some point in the past. I used an Apple device, and I am sure that the data needed was also drawn from the massive dataset of other apple users and devices all over the world.
The frustrating aspect of the activity was the actual words being predicted. I found it nearly impossible to guide the sentence in the say that I wanted it to go, and it made me feel a bit powerless, not having any agency over the process. Although I was given some choice in which words to select, I wished that there was a more diverse set of words to choose from as I typed. I wanted to type more, but I felt like what I had already down was so non-sensical that it would be counterproductive to continue.
I don’t remember seeing sentences like this in any type of serious publication, perhaps on a blog but that’s it. I don’t feel that I would ever express myself in this way because the tone and word usage doesn’t fit my voice at all. I do wonder if the algorithm should have done a better job with capturing my voice though. After all, I own multiple apple devices, all using the same account and have done so for a number of years. I would have hoped that with all of the data that it has collected on my typing , it should have made the device a bit “smarter” and provided predictive text more closely resembling my particular voice.
I am curious about the potential for this type of predictive text algorithm in education. If it got to the point where it was a bit more adaptive and more accurate with understanding the context of what the writer is trying to say, perhaps it could be used to assist struggling writers in the classroom. I could see some type of software that helps students generate writing ideas based on a starting sentence, or provides suggestions and assistance to struggling writers to help them stay on track and be less frustrated as they write.
Hi Brian,
I really liked the video that your produced with the software that you used and showing the process of the predicative text creating your characters. I also understand your frustration of not being able to guide it or help it a long with real ideas. In my experience I noticed that there was a bit of a overlapping of what I had said about COVID to people and what the predictive text was throwing out as I typed it on my phone, which evidently has memorized some of the things I had typed on it. I like how you can see how this can be used in creative writing in the future.
Hi Alexandra,
Thanks for the comment. Producing the video was actually easier than you would think. There wasn’t really any fancy software, I just did a recording of my iPad screen using this easy method (https://www.cnet.com/how-to/quickly-record-a-video-of-your-iphone-or-ipads-screen-with-just-a-couple-of-taps/). Then, I did a quick edit in iMovie and uploaded it to YouTube. The iMovie step wasn’t really necessary though, and I could have saved some time and just sent it straight to YouTube (just wanted to add a cool transition!).
Hi Brian, I liked the video you have created and I have never heard of this tweetgen before. I will check it out. I liked your question regarding this in education, on my end I said I can see this being a good idea for those students who struggle with stories and using their imagination. I thought this would be a fun story starter task for the students.