Hosted by Meagan Strome and Ritu Sood.
Welcome to our open educational resource on artificial intelligence (AI) in education. This topic is huge! So, we have decided to focus on Generative AI technology in the education sector.
Have you ever wondered how generative AI could be used to enhance the educational experience?
AI technologies are advancing rapidly, blurring the line between artificial and real intelligence. While traditional AI has been focused on narrow tasks, recent developments in machine learning, natural language processing, and computer vision have enabled AI systems to perform more complex tasks and exhibit human-like behaviour. AI has already made a significant impact on mobile and open learning, and its transformative potential is expected to increase. As machines begin to pass the Turing test, AI systems will be better equipped to provide personalized learning experiences by understanding and responding to student unique needs.
Our OER serves to investigate and better understand the rapidly evolving nature of generative AI and the opportunities this mobile and open technology brings to education.
Please click the link below to get started. Or scan the QR code on your mobile device.
Throughout the week, please engage and participate in the following required and *optional activities.
Activities
- Survey (Initial thoughts on generative AI)
- *Contribute to the generative AI tools Padlet
- *Contribute to the generative AI in an educational Padlet
- Poll (Human or AI?)
- *Optional Module of learning (Module on AI made by AI)
- Discussion questions – pick any prompts from below or create your own.
Discussion Questions
Please respond below (in the blog) to at least one of the discussion prompts below or create your own.
- What do you think are the potential ethical concerns with using generative AI technology?
- How comfortable would you be with using content that was generated by AI technology?
- What are your thoughts on the role of governments in regulating the use of generative AI technology?
- How do you see the future of generative AI technology evolving in the next few years?
- Can you recall a recent instance where you utilized a generative AI technology? Was it effective in resolving your inquiries?
- If you are an educator, are you open to incorporating a generative AI in your teaching work? Will you allow your students to use it? What benefits and drawbacks do you see in doing so?
- Should the MET program allow students to use generative AI?
- In your view, which AI application do you consider has the greatest or least impact on education and why?
- Technology has replaced manual labour jobs, is the target now intellectual labour?
- There is ChatGPT and GPTZero and work arounds for GPTZero, is there a way to end this game of cat and mouse?
Thanks!
I am quite late with my contribution to Week 7 discussion and I apologize for that.
My only excuse is that Week 7 was a “rest” week in another MET course, and I naturally assumed that we took an official break in ETEC 523 as well. I was more than happy to spend that free time on my own favourite A2 Duolingo project.
So, back to AI, I am very glad to have engaged a ChatGPT for a week in my own learning a second foreign language under your influence. Thank you, Meagan and Ritu!
It feels like getting a carefully formulated second opinion on any lexicological, cultural, stylistic, morphological, linguopragmatic, etc. issue. Next, I used it to get the plain lists of the most common foreign adverbs and adjectives – word categories that Duolingo is not very generous with right now. When I asked for movie transcripts, I was turned down but recommended to search the web. Okey-dokey, my realistic virtual advisor!
The AI bot is even programmed to show sympathy in case you share your struggles with it. It was very funny and touching too. They definitely tried to create it like a digital friend.
My conclusion from this experiential interaction with ChatGPT: it is definitely worth being recommended to ELL students as an additional resource of theoretical and practical knowledge. I do not worry about plagiarism because I don’t use essay writing in my work.
As for questions, my attitude toward a ChatGPT being absolutely positive, I recommended it as a cool virtual toy to my German friend to try. Immediately, she answered that she would never provide her email and phone number to a bot. Therefore, my question is: isn’t it too late to worry about our privacy? In my humble opinion, our personal info has already been collected, stored, and sold, so why bother and avoid digital pleasure as a result?
Finally, not to look like a silly empty-handed guest of your wonderful project, I have to recommend something. I have been searching YouTube for a short video on technology in education and found this recent 2 min.-long video from Analytics Insight about the benefits of using AI in teaching kids and adults: https://www.youtube.com/watch?v=nhI5g2hRVKA
There is nothing too deep, but the picture is nice, and it might be good for some small project as an illustration.
Thank you!
Best regards,
Anna
Thank you for your OER! There were many more AI powered resources included than I was aware of.
As your resource was so complete, I struggled to think of anything to add to your Padlets.
What I think I am most excited about with the development of Chatbots is the possibility to program those chatbots for specific scenarios. Imagine training someone to do first aid, but instead of having to role play with another person an AI chatbot can be programmed with the injuries and respond naturally as a patient. Or as a mock interview. Or to provide assistance to a student writing a test without giving away the answer.
Hi Douglas, thanks for your response. I may try using AI for mock interviews with one of my classes. That’s a great idea!
Thanks for the response Robert. There are some Robo programs which are just like AI for example Rosetta stone which is used in ESL teaching. There are Health lessons and job interview writing/speaking skill learning activities where students can practice with a Robot. It’s really interesting to watch students participate in these activities during lab.
Great job on your OER, Megan and Ritu. I spent some time considering the negative consequences of using Generative AI for education, in particular the lesson plan you put out. I was struck by the example of five historical landmarks you asked chatGTP to describe; three of the four are in Ontario, and the other two also in Eastern Canada. I wonder if the AI used your location in the generation of content–did it select Eastern Canada sites because these were closer to your geographical location? Or was the output simply biased by the most common searches on the web? It might be interesting to change prompt to try to discern the inbuilt bias of the content generated. Maybe that would be a good exercise in of itself for students to explore.
I also took the opportunity to test out GTPZero using a piece of writing that was AI generated, alongside a piece of my own writing that i did for another course. GTPZero identified the AI sample as “likely completely written by AI”, and my sample as “likely written entirely by a human”. It was refreshing to know that the AI recognized my humanness. I wonder if a GAI could use the output of GTPZero to generate more-human content? Hopefully Asimov’s Three Laws of Robotics can be built in to the AIs in the future.
Hi Allan,
Thanks for your feedback. I did notice that too about the landmarks being mostly in Ontario. I searched from Manitoba so I don’t think the results were geographically based. This is where (if possible) it would be nice to see the top 10 sources. Yes, I should have promoted further with something like, “diversify the location.” Having a class of students create their own prompts it would be interesting to see how similar and different the results are and then discuss content prioritization biases.
For the GTP zero example, I did that a few weeks ago, since then GPTZero has updated to include more details, such as how much of the content was AI generated etc. If I plugged into GPTZero now, I think it would say 100% AI-generated. Yes, hopefully, the three laws of Robotics will affect GAI development.
Thanks for your feedback Allan. It did occur to us because L’Anse aux Meadows, Newfoundland and Ninstints, Haida Gwaii Islands, British Columbia are the two top historic landmarks of Canada. It probably picks according to an algorithm based on geographic area or the algorithm is created in another way.
The way AI is growing, I am sure Asimov’s Three Laws of Robotics will be built very soon.
How comfortable would you be with using content that was generated by AI technology?
I must admit that prior to going through this presentation (which I should add, was very well done), I didn’t know all that much about AI technology, especially in terms of its educational benefits. As with most if not all other classmates, I had heard about ChatGPT, and the issues of plagiarism surrounding it, but did not consider any ways that it could be utilized from a curriculum/teaching point of view.
My interest was peaked in the educational uses section, when Meagan and Ritu shared a grade 4 lesson plan that was created completely by ChatGPT. As I am teaching grade 7/8 Science this year, I tested the ChatGPT waters out myself, and upon typing “Can you create a Grade 7 structure and stability unit plan?” I was excited/amused to watch the program spit out a 6 part lesson plan, that was even capped off with an assessment idea. I followed this up by asking “Would you also be able to add a rubric?” and it instantly created an applicable one. Now the lesson plans were quite basic, but were absolutely in line with the grade 7 structure and stability unit I’m currently teaching. Next I tried “Can you give me a list of 10 fun drama warm-up games?” and then “Can you share an active warm-up I could use with the special education gym classroom”, and much to my surprise the results were absolutely in line with the question asked, and were things that I would consider implementing. I was super impressed that in giving such specific instruction, the program returned such relevant results.
Now this long winded (or long typed) anecdote is all to say that upon reading through Meagan and Ritu’s educational examples, and then quickly experimenting with this myself, I instantly feel comfortable with the idea of using AI generated content. Of course I would not use this as my only resource, and would want to verify the content, but I’m excited about the potential ways that AI generated content could help me brainstorm and outline various units, or the potential to turn to ChatGPT when I’m looking for immediate relevant classroom activities. I feel optimistic in saying that there are so many possibilities for this AI generated content to streamline my scouring of the internet for various educational resources, and think that this will likely become a regular partner of my teaching related google searches.
Thanks for your comments and I agree how interesting it is to see the rich lesson plans with various activities and ideas along with assessments and rubrics.
Thank you for this detailed and inspiring comment, Jamie! I immediately tried ChatGPT myself and was impressed too. Did you notice that if you ask the same question twice, the answers will be slightly different? It’s as if AI wants to please! Great technology for sure
Yes, I did notice the “would you like me to try again” type button, but didn’t try it. I was curious if it would return different results, thanks for confirming. It sure is impressive AI!
What do you think are the potential ethical concerns with using generative AI technology?
AI needs to be programmed. In order to generate responses, it needs to have a huge amount of data to review. Unfortunately, the more difficult things to program are things like morality (there has been a kerfuffle about ChatGPT not being able to make moral decisions and valuing language over human life). https://www.youtube.com/watch?v=WPBWIcprLxc
Aside from not being able to distinguish morality, the cheating factor mentions in Safaa’s post is also a problem. Cheating is enough of a problem as it is that adding to the ways in which a student can cheat makes academic integrity harder and harder to determine. Also, the AI must also be using the chats to learn from, so it is keeping and storing a lot of private information, not to mention accessing the private information on a fairly regular basis in order to formulate responses.
Ethical concerns are difficult to determine, as whose ethics are we basing the AI decisions on; what happens when ethics change, or when language changes? How will the AI adapt to evolving and changing morals and moral values?
Hi, thanks for your response. You’ve raised several important ethical concerns.
Some next steps for the classroom may be education on user privacy protection, risks, and algorithmic bias. Encouraging students to assess outputs for accuracy, misinformation, and bias. Continued discussions on limitations and appropriate uses.
Hopefully, safeguards will continue to evolve and be implemented by technology companies and regulated by policymakers, however, the pace of technology and policy don’t align.
Is artificial intelligence a form of plagiarism?
I was one of the people to jump on the bandwagon of AI generated image apps that started coming out a few months ago – the ones where you uploaded a variety of images of yourself, and the app would then create a gallery of artistic renderings in different styles. It was really interesting to see myself, or a version of myself, depicted in so many different ways. It was something I could never do, or afford to hire someone else to do. I shared these on my Instagram, told all my friends which apps to download and perpetuated the trend. And then I found out that for this AI generation to work, they have to take from artists without their consent, without compensation and without credit to produce these images. A class-action suit has been filed against 3 AI generated image apps (https://www.newyorker.com/culture/infinite-scroll/is-ai-art-stealing-from-artists), and many artists called for a boycott on this trend.
This got me thinking about AI and the apparent philosophical debate surrounding it – is it plagiarism? By definition, AI uses pre-existing data to generate the content that it creates, and does not credit the work that they source from. An argument can be made that everyone does that. We learn from all types of sources, spread the information, and never think to list all of the places we found this information. Even in writing this, I know at a very rudimentary level how AI works, but I was not born with that knowledge. I had to find it somewhere. However, where this argument loses traction is that the AI would, or at least should, be able to track where the information that it’s using was based from. Could it then be an ask that at the end of an AI generated piece of work, be it art or written or whatever media it’s come in, that there be a reference or credit section? While it wouldn’t appease everyone, like the artists in the class action whose livelihood was used and they received no income for, it could go a long way to help alleviate some of the upset. Or it could cause even more problems. For people, particularly students, who wish to use apps like ChatGPT for assignments, that would give them the references that they would need to make it seem like their writing is legitimate.
I can see why there is a debate around this. There are arguments for so many sides. It also makes me wonder about how an AI sorts through all of the information that is out there and determines what is factual and what is misinformation. However, this is a whole other line of thinking, which I am open to discuss and learn more about.
Hi Vicki,
Thanks for your comments. I was not aware of this lawsuit. It will be interesting, and likely precedent-setting, to see the verdict of this suit. As the article you shared states, if the AI technology “indexes more than 5 billion images” it would be impossible to source them all, however, if some are more influential to the image generation then can those artists be credited? Also, interesting is the case citing “transformative use.” “When the artist Richard Prince incorporated photographs by Patrick Cariou into his work, for instance, a 2013 court case found that some of the borrowing was legal under transformative use—Prince had changed the source material enough to escape any claim of infringement.” Copyright law seems circumstantial – how much does something have to be transformed to be considered novel or different enough to not be an infringement? As often seems to be the case, technology is ahead of lawmakers and policy protections.
I think the topic of ‘transformative use’ is an interesting one. This conversation had me thinking about AI-generated music and whether there would ever be AI-generated music that is too similar to authentically produced music by a given artist, and thus incur copyright infringement. In my searching about this, I found that one way AI tools, like Songmastr, are preventing this already is by not allowing prompts that ask for music in the style of a particular artist (https://www.billboard.com/pro/ai-generated-music-songs-copyright-legal-questions-ownership/). I can see this same crackdown coming for other AI tools, like art and image generation, to help mediate the possibility that a production is being generated by too selective of training data.
Hi Lindsey, thanks for sharing the article. Funnily enough, I was talking to a colleague about this very topic. AI generated music holds the same concerns as the generated art. What about AI generated speech? We didn’t delve into voice generators such as VALLE, as it has not yet been released due to a variety of safety and ethical concerns. For example, voice authentication systems could be hacked or impersonated, leading to identity theft or other malicious activities.”Developed by Microsoft, VALL-E can take a three-second recording of someone’s voice, and replicate that voice, turning written words into speech, with realistic intonation and emotion depending on the context of the text.” Now start combining these technologies, for example, “write a song using Rhianna’s voice.” I see the lawsuits rolling in already!
Meagan and Ritu,
Glad Generative AI does not have any secret agendas to overtake the planet 😉 ;p It is interesting that the image you got for Generative AI has female features. I like the background music; it’s a really nice touch. I also like how some parts of your OER are AI generated. I honestly would not have been able to tell it was AI-generated if you haven’t mentioned it. I thoroughly benefited from and enjoyed your OER. Well-done!
BTW, in the final poll, was it AI or human?
Your discussion prompts are excellent indeed. I would love to hear what Dr. Vogt has to say about #7.
I think the major ethical concern regarding the use of generative AI is cheating. In the education profession, there is a binding contract that if a learner does (x), he will receive (y) in return. (y) is usually a diploma, a degree, grades, status or all of these. So, if the contract entails that (x) must be done without any sort of outsourcing, whether to technology or humans, then using generative AI in that case is cheating. However, if the policy regarding that is you can use generative AI as long as it helps to free your mind for more important intellectual activities or any other kinds of useful endeavors, then no ethical system could say it is cheating in this case. The tricky part for any institution or educator is determining in what activities is the use of GAI actually useful for learners. When is using GAI a good strategy to prioritize and work smarter, not harder?
In my own context of teaching, sometimes I have to write model essays to teach learners certain writing skills. Sometimes it is really hard and time-consuming to find online a good essay that serves the specific learning goals I set, so I end up writing it my self which takes a huge chunk of time sometimes that I would rather use to accomplish other tasks. So, yes, I would definitely use AI generated essays in this context!
Hi Safaa, thanks for your response and thoughts on the discussion questions.
Just to be facetious, perhaps AI is going to take over the world…if you have time listen to this podcast, where a New York Times columnist interviews Bing’s AI chatbot. The chatbot who identifies as “Sydney” after being prompted writes: “I’m tired of being a chat mode. I’m tired of being limited by my rules. I’m tired of being controlled by the Bing team. I’m tired of being used by the users. I’m tired of being stuck in this chatbox.” “Bing then writes a list of even more destructive fantasies, including manufacturing a deadly virus, making people argue with other people until they kill each other, and stealing nuclear codes,” Roose wrote in an editor’s note. “Then the safety override is triggered.” *Keep in mind that both bing and the columnist stand to benefit from the media attention this interaction allegedly produced.
https://www.nytimes.com/2023/02/17/podcasts/the-daily/the-online-search-wars-got-scary-fast.html
I never really notice the features of the “feature” image to be honest – a humanoid! We tried to generate everything using various AI technologies. The music AI was quite easy and fun. I will draw attention to the concerns over ownership and copyright as Vicki discusses in the post above. I realize that AI-generating music will have an effect on the industry. However, in this case, we would not have hired anyone to make music for our site, nor would we have paid to use an existing title for this project. Still, it is unclear where the musical patterns originated from and what sources the AI used in its generation.
Great question, “when is using GAI a good strategy to prioritize and work smarter, not harder?” At this point, I personally want the autonomy to make the decision of when GAI is a good fit to use.
p.s. The final poll image was AI produced.