Click on the imageThe first difference I noticed between Melissa and I’s Task 3, is that we used different voice recording software. Melissa used Google Docs speech-to-text function while I used Speechnotes.
I found it very interesting when Melissa mentioned Google’s AI process to find the most suitable entries for word prediction. It made me question if all text-to-speech software uses AI to find the ‘best guess’ words. If so, is it Google’s access to all of our searches that makes it more accurate as a speech prediction software? Does Google speech-to- text work as accurately for people who have accents?
Although Melissa’s Task 3 was written a few weeks back, it puts me in mind of the podcast we just listened to by McRaney (n.d.). According to McRaney’s (n.d.) podcast, an algorithm guessed what words Melissa meant to use, by pulling from it’s database of words that had been accessed frequently before. When Melissa entered some unfamiliar data, the story changed. When Melissa speaks the names of Jack Woody and Ian Watt, perhaps Googled less than their counterparts; Eric Havelock and Walter Ong, the algorithm was at a loss for what to do. The question that comes to mind from this event is: when we find what we are looking for, should we stop looking? Has our lack of adventure caused our monoculture (Rumsey, 2017)?
Both Melissa and I, in our Task 3 entry, delve into how the lack of punctuation in our transcribed texts has affected their meanings. I discuss how the spatial quality of the written text is not a true rendition of the spoken (Hass, 2013, p.9), making it a broken translation. Melissa also discusses the spatial quality of her written text as being: “…one big and heavy block of text” (Drake, 2021). I wonder if the conversion from written language to hypertext bypasses this heaviness, as hyperlinks chunk useful information, making texts lighter.
Both of us use a doorway to introduce our voice stories. While Melissa uses a quote and a wordle butterfly of the words in her story, mine uses an image of the Sierra Nevada de Santa Marta, the location where the story took place. This might seem like a small detail when comparing our two tasks, but to tell you the truth, I did wrestle with whether or not to include a picture. I wondered if I was changing the reader’s experience by including the photo. Was I scaffolding them though the jumbled words of my recording? In the end, I chose to include it because I couldn’t stand how much written text there was on the page. In contrast, Melissa introduced her story with a quote and a wordle butterfly, one of the characters in her story, with vocabulary from her story, creating a ‘pleasing tension’ (Bolter, 2001, p.63) and allowing “…the viewer to consider the words as images or abstract shapes rather than signs” (Bolter, 2001, p.63).
Bolter, J. D. (2001). Writing space : Computers, hypertext, and the remediation of print. Lawrence Erlbaum Associates.
Haas, C. (2013). “The Technology Question.” In Writing technology: Studies on the materiality of literacy. Routledge. (pp. 3-23).
McRaney, D. (n.d.). Machine Bias (rebroadcast). In You Are Not So Smart. Retrieved from https://soundcloud.com/youarenotsosmart/140-machine-bias-rebroadcast
Brown University. (2017, July 11). Abby Smith Rumsey: “Digital Memory: What Can We Afford to Lose?” [Video] YouTube. https://youtu.be/FBrahqg9ZMc