When I was in high school in the late 1990s, cell phones were just becoming ubiquitous among my Generation X peer group. However, the way we used these early generation phones for communication purposes did not look exactly like how most people use them now. Text messaging, in particular, only became realistically functional around 1999 when the feature started working across different networks, allowing one cell phone user to text any other, regardless of whether their service provider was AT&T, Sprint, etc. Yet even with its lower-than-calling cost, the new medium of texting did not catch on right away. I can distinctly remember sharing a sentiment with many others at the time: why would I text when I could just call? Speaking live to the other person just seemed so much easier, so much more efficient and timely, so much more…familiar. Phones– whether in a glass booth on a city street corner, attached to the wall of your house, or in your pocket–were for calling and talking.
For a while anyway, neither the early cell phone user nor the technology hardware was quite ready to usher in a significant replacement of live verbal over-the-phone communication in favor of finger-pecked messages. The users at the time, still getting used to digital communications in the form of electronic mail (then electronic-mail, then e-mail, and finally email), had spent their entire lives understanding landline phones as talking and listening devices. Why text from a phone what could be much more easily typed and emailed from a computer? The cell phones of the 1990s and early 2000s, without full alphabetical keypads, made texting a very clunky process.
Despite these challenges, the utility of the text message via cell phone ensured its survival, and as Blackberry phones introduced full keypads (2002) and the iphone introduced a full touchscreen keypad (2008), the user experience for texting improved dramatically. By 2011, Pew reported that almost a third of cell phone users who used text messaging preferred receiving and sending texts over calls. A little over a decade later, texting achieved status as the preferred choice of communication over calls, in-person conversations, video calls, and email. Writing (of a sort) actually defeated speaking.
The matter seems to be settled then, for now at least: texting is generally preferred as the means of communication between people today. But what about how humans interact with computers?
Virtual assistants in the form of Alexa speakers and Apple’s Siri have been around for over a decade, and they are technologies that exclusively work via voice activation and verbal inputs from the user. Speech-to-text programs have also been around for quite some time, and their advancement has made it easier for many people to dictate text messages, “write” emails, and utilize word processing programs like Google docs and Word. For many, typing textual inputs for their computers and devices is not a preferred method over speaking into the devices. The path to this point was/is not without challenges, however:
The latest generative AI models like ChatGPT can also respond to text or speech inputs from users. Considering the rapid integration of generative AI into existing technologies like phones and personal computers, how people choose to interact with the generative AI will largely drive both how educators might train people to use the AI and how the developers of generative AI adapt the user experience of the products.
As an instructor of writing, I decided a couple of years ago that generative AI might not be such a terrible disruption to my line of work –IF– the interface of ChatGPT and others like it was actually forcing the users to write their prompts for the machines. In fact, like others, I saw prompt writing as a new and exciting way to teach both basic writing skills and critical thinking. After all, getting a generative AI model to do what the user wants relies heavily on clear written communication (along with persistence and revision and creativity). However, as these chatbots increasingly incorporate better speech-to-text capabilities, I wonder if the idea of prompt writing will soon lose out in favor of prompt speaking—similar to how a reverse situation unfolded with verbal cell phone conversations losing the battle against written text messaging. If science fiction is any indication of what the future of human-computer interaction will look like (and sci-fi has quite the impressive track record), then keyboards could be on their way out. Hal and JARVIS and Kit have superior listening comprehension.
Already, some computer coders have turned away from natural language text prompts of generative AI models in favor what they are calling “vibe coding,” or speaking their instructions rather than typing them. This does not seem like a good sign for my plans to teach writing through the lens of AI prompting, but we will just have to wait and see if speech prompting catches on and influences the way that the tech industry updates AI experiences and interfaces for users.