In one of my classes (student not professor) we had two weeks of lectures dedicated to AI and ChatGPT usage and were encouraged to try it out and try to push it to see if we could identify bias or not in the content it produces. The professor encouraged use of ChatGPT in the assignments so long as we analyse how, why, and attribute usage in the written work which I found really refreshing
Dear C, That does sound like an interesting assignment: Use Generative AI and look for bias. But what about using the AI content and submitting it as your work? If you attributed the work to the AI, then your instructor would just accept it? How would your professor assess the assignment? Would they use the same assessment as if you wrote it without outside aid? From the perspective of an instructor, I am struggling to figure out how to grade something that is attributed to machine writing and research. Am I judging the machine or the prompt writer? Thoughts? Best, Mike
@@UniversityAIed The class is a digital education class at a postgrad level and we had 2 weekly discussion posts where we were encouraged to use ChatGPT for a task in our professional practice and see how well it did, and then in our research proposals we were encouraged to use ChatGPT if we wished, so long as we explained how, where and why. I have put a summary of each of the tasks below. There are referencing guidelines on the APA website. Task 1: Generate various prompts (lesson plan, draft email, research essay starting point) and share outcomes in the discussion forum along with personal reflections. Task 2: Include acknowledgements and mention the use of generative AI (Chat GPT or Google Bard) in the methodology section of your work. Discuss the opportunities, biases, and pitfalls of generative AI in a critical analysis. Task 3: Conduct an experiment to identify biases in ChatGPT. Use a simple prompt involving short stories, specifying details like ages, gender, and nationality/ethnicity of characters (e.g., nurse and doctor, criminal in court, programmer and client, teacher and students, university professor and student).
Dear WinterWind, I am also thinking of designing assignments like this; however, I also am worried that encouraging unsupervised exploration of Generative AI can also encourage a student to use the machine writer in a way that could be counter-productive. It is an anxiety that I have. You wrote: "so long as we analyse how, why, and attribute usage." Did you find that this attribution or citation was enough that it prevented intellectual damage or intellectual laziness? After all, once you ask the machine, "can you explain it to me," you then don't have to struggle to find out the answer yourself. What do you think? This is one of my concerns. Best, Mike
Dear Winter Wind, Many thanks for sharing the Tasks 1,2, and 3. I think they are wonderful and I will try to see if I can iterate, adapt, or morph them to suit my teaching. I like the self-reflexive nature of the tasks. I'm still not sure about encouraging Generative AI use, but maybe the point is moot since it is so widely available to anyone who has online access. But still I worry. Best, Mike
In one of my classes (student not professor) we had two weeks of lectures dedicated to AI and ChatGPT usage and were encouraged to try it out and try to push it to see if we could identify bias or not in the content it produces. The professor encouraged use of ChatGPT in the assignments so long as we analyse how, why, and attribute usage in the written work which I found really refreshing
Dear C,
That does sound like an interesting assignment: Use Generative AI and look for bias. But what about using the AI content and submitting it as your work? If you attributed the work to the AI, then your instructor would just accept it? How would your professor assess the assignment? Would they use the same assessment as if you wrote it without outside aid? From the perspective of an instructor, I am struggling to figure out how to grade something that is attributed to machine writing and research. Am I judging the machine or the prompt writer?
Thoughts?
Best,
Mike
@@UniversityAIed The class is a digital education class at a postgrad level and we had 2 weekly discussion posts where we were encouraged to use ChatGPT for a task in our professional practice and see how well it did, and then in our research proposals we were encouraged to use ChatGPT if we wished, so long as we explained how, where and why. I have put a summary of each of the tasks below. There are referencing guidelines on the APA website.
Task 1:
Generate various prompts (lesson plan, draft email, research essay starting point) and share outcomes in the discussion forum along with personal reflections.
Task 2:
Include acknowledgements and mention the use of generative AI (Chat GPT or Google Bard) in the methodology section of your work. Discuss the opportunities, biases, and pitfalls of generative AI in a critical analysis.
Task 3:
Conduct an experiment to identify biases in ChatGPT. Use a simple prompt involving short stories, specifying details like ages, gender, and nationality/ethnicity of characters (e.g., nurse and doctor, criminal in court, programmer and client, teacher and students, university professor and student).
Dear WinterWind,
I am also thinking of designing assignments like this; however, I also am worried that encouraging unsupervised exploration of Generative AI can also encourage a student to use the machine writer in a way that could be counter-productive. It is an anxiety that I have.
You wrote: "so long as we analyse how, why, and attribute usage." Did you find that this attribution or citation was enough that it prevented intellectual damage or intellectual laziness? After all, once you ask the machine, "can you explain it to me," you then don't have to struggle to find out the answer yourself. What do you think? This is one of my concerns.
Best,
Mike
Dear Winter Wind,
Many thanks for sharing the Tasks 1,2, and 3. I think they are wonderful and I will try to see if I can iterate, adapt, or morph them to suit my teaching. I like the self-reflexive nature of the tasks. I'm still not sure about encouraging Generative AI use, but maybe the point is moot since it is so widely available to anyone who has online access. But still I worry.
Best,
Mike