Learning never stops; this now includes both humans and Artificial Intelligence. As I type this blog post, I find myself either tabbing to accept the suggested word or ignoring the suggestion. Being prompted to type what auto-text thinks I should be writing can be annoying and, if I am not careful, I end up writing a word that I did not mean to write or, worse yet, pressing ‘send’ on a message or email with one or two unintended words. Although I appreciate its usefulness on some occasions, it irks me when I am given the wrong suggestion, as in the case of Grammarly’s use of double commas on a salutation (since when did adding a comma after ‘Hello’ or ‘Hi’ and before someone’s name become the grammar norm?)
However, no matter how annoying auto text can be, it is harmless compared to the new AI generated text applications in existence. AI has evolved, making it more tempting for people to rely on its suggestions, and, as a result, lose the ability to learn, if they are not careful. For teachers, AI is the new distractor. AI is also the new plagiarizer. Students (like everyone else) are lured by the easiness of having a typing assistant, an idea generator, and a grammar checker all in one.
I am not including Google Translate in this group, for I think that most times it is used to translate original produced text (in other words, student generated ideas). Google Translate would be what Vygotsky calls a mediator – the means for a student to reach the zone of proximal development.
Of course, my intention is not to cause alarm, an Orson Welles’ “War of the Worlds’ pandemonium moment. Regardless of all AI’s negative denotation and connotation, we can learn to use it to our advantage without letting it take away our confidence or give us a false sense of one. Not all is lost.
What should we do as teachers?
Here are three ideas I have in mind for next year to build AI into my lesson
- Low-stake competitions: For example, students could compete against AI by authoring their own stories based on a preliminary idea that I provide. They would then compare their self-generated text with the AI version. I am thinking that, to keep it fair, the text should be assessed based on a rubric (student generated, for example).
- Students as copy editors: Another idea is for students to revise, edit, and proofread the AI generated text. They could add transitions, find the off-topic idea, or check for fallacies.
- Ethical discussions: I also plan to include articles on the topic to discuss issues of ethic and plagiarism. For example, AI can generate false references which can be easily checked for authenticity (Sharples, 2022). This would be an effective way to discuss academic honesty.
Trying It Out
If you are interested, you can play around with AI generated text to get a feeling of what it looks like. For my example below, and to put AI to the test, I used the demo version of InferKiti. The first sentence text is mine. The italicized text is AI-generated.
Driving a driverless car has its advantages.
If one driverless vehicle is going to become the norm, the cars themselves will need to be able to communicate quickly with one another. That will make communicating with real people all the more critical.
Sharing details such as what route the vehicle has in mind, whether it is able to drive any further, and when it will be able to return home could greatly improve safety.
Driving without a passenger is going to be a nightmare for all of us.
The technology will also make more efficient use of public space.
(As in most AI text generated text, you can also input words that you wish AI to include in the generated final product).
Food for thought
As we moved to the future of AI, Issac Azimov’s three rules of robotics come to mind:
- A robot should do no harm
- A robot must follow orders unless these override the first rule
- A robot must protect its existence unless to do so it breaks the first and second rule.
We all know plagiarism is a no-no. There go rules 1 and 2. Any thoughts?
References
Sharples, M. (2022, May 17). New AI tools that can write student essays require educators to rethink teaching and assessments. LSE Impact Blog. https://blogs.lse.ac.uk/impactofsocialsciences/2022/05/17/new-ai-tools-that-can-write-student-essays-require-educators-to-rethink-teaching-and-assessment/
Further reading (If you come across any, please share):
Marche, S. (2022, December 6). The college essay is dead. The Atlantic. https://www.theatlantic.com/technology/archive/2022/12/chatgpt-ai-writing-college-student-essays/672371/ (Note that viewing might require a subscription)
Monge, J.C. (2022, December 12). Canva’s magic write – An AI-powered writing assistance. Medium.com. https://medium.com/technology-hits/canvas-magicwrite-an-ai-powered-writing-assistant-b4801bbfdf38
Rikab, W. (2022, December 10). How to spot if the article you’re reading was written by an AI. Medium.com. https://medium.com/@waleedrikab/how-to-spot-if-the-article-youre-reading-was-written-by-an-ai-8557f75de32
*Link to Orson Welles’ “War of the Worlds” article: https://www.smithsonianmag.com/history/infamous-war-worlds-radio-broadcast-was-magnificent-fluke-180955180/
*Link to Asimov’s three rule of robotics: https://www.britannica.com/topic/Three-Laws-of-Robotics
Cecilia Aponte-de-Hanna
Hi, my name is Cecilia. I love taking part in good brain awakening discussions. Blogging, I find, lends itself for that. I also believe in sharing my skills through scholarly practice, which is why I write regularly and have presented at several conferences, including TESL Ontario, TESL Toronto, CALL, and at Seneca College. My M.A. in applied linguistics along with my skills and experience have led me to my current position at Centennial College, where I teach English and ESL in the School of English and Liberal Studies. I'm truly passionate about what I do: teaching, writing, creative expression, and helping my students (both L1 and L2) gain agency and take control of their own learning. Thank you for your readership and I look forward to reading and answering your comments. You can find me on Twitter @capontedehanna
There is so much more to explore regarding this subject, including all other types of AI generators out there, which wouldn’t fit in one blog post . . . Some require a subscription such as ChatGPT, which is one of the applications available through OpenAI. If you are interested, you can read more about this team’s research initiatives on their “About” page. If you think H5P is an amazing tool (which it is), you’ll also find that this company’s work is also packed with amazing possibilities!
Cecilia,
I think the best way to vaccinate your students from AI is to have them complete all writing tasks with a pencil on paper while you are in the room. Other than that, we have lost control as there are so many digital supports already within their reach. My future shock reference here is the movie Westworld. https://tinyurl.com/mrxawmy9 Robots just can’t be trusted! 🙂 This is an interesting and important topic for our profession. Thanks for kicking it off.
John
Hi John, If only pen and paper were the only option! However, in our digital world, we cannot get away from typing and obviously, computers (like I’m doing now).
Perhaps AI is one way to drive innovation.: check what AI has to say and come up with something completely different. The skill of critical thinking can prevail if we want to be in control instead of being controlled . . .
Following up with your movie suggestion, I’ll check it out, but it looks like it could be equally shocking!
Cecilia
Thank you, Cecilia, this is an incredibly relevant topic, and as my colleague I’m grateful you’ve investigated some possibilities in the virtual unknown. Admittedly, I’m a digital immigrant. It’s becoming incredibly hard to communicate to our students the relevance of “authentic voice” in the context of post-secondary online writing that is graded. I like your suggestions about building in student agency, but how do we, and conversely our students, discern what’s reliable information courtesy of AI? I recently sampled a few queries in ChatGPT and the result felt like a Wikipedia experience except the “contributors’ identities” had been scrubbed. It was as if ChatGPT was a master plagiarizer. As educators, I hope we find ways to show students that relying on AI to reproduce unoriginal content makes students steal from their innate ability to learn for themselves.
Perhaps we need to input the prompts we ask students to write about in these chatbots. I’ve entered some which the generator is unable to answer. But beware: the new AI Chrome extensions take away the need for self expression … although the intention still lies on the person’s. The new capital is the prompt not the message, it seems.
Cecilia,
Interesting thoughts.
What I think is really amusing (ironic?) about the situation is that this capability seems to have come to fruition at about the same time as the ‘Zoom era’ of education has supposedly ushered in a new era of unlimited flexibility and choice in higher education (as an administrator would no doubt put it!). Surely the concern about ChatGPT would be nowhere near the level it is at now had this not become the reality. So, it is almost as if technology has undermined itself, or at least certain near-utopian visions of it.
Your timely comment makes me think of a Margaret Atwood’s dystopia and her apt words “A Word After a Word After a Word is Power”. Do we relinquish our power and become voiceless (or more voiceless) or do we use this moment to think, reflect, and speak? There is no utopia but we shouldn’t surrender our mental capabilities.