One of the biggest worries about generative AI is in terms of how it could impact the development of the ability to write and the learning that is, in essence, formed and evolved through the construction of sentences, paragraphs and the outputs of writing from songs to blogs to academic essays. There’s been some really thoughtful work in this area aleady and Anna Mills has collected some amazing resources that offer a range of perspectives and approaches as well as plenty of food for thought about impacts and issues. This series of videos ‘Generative AI practicals’ is designed to suggest ways in which tools like ChatGPT and Google Bard might be used by academic staff and students in ways other than pumping out text indiscriminately and uncritically! In this video I isolate one element from a marking rubric and using two genAI tools ask them to assess a paragraph and then suggest alternatives across grade bands.
This video looks at how students might start to think about bouncing ideas around in ChatGPT (or similar) to get to better and/ or deeper understandings of the subjects they are studying. This is not about producing content but engaging with ideas.
There are loads of things we (in HE and education more broadly) need to think about and do when it comes to generative AI, both cognitively and practically. I am alert to and concerned about the ethical and practical implications of generative AI tools but here want to focus on ways in which we (teachers and students) might find ways to use these tools productively (as well as ethically and with integrity). My view is that the ‘wow’ (or OMG) moment experienced when you witness tools like chatGPT spouting text needs to be looked beyond and ways in which the mooted idea of AI personal assistants can actually be realised need to be explored and shared. As a compulsive fiddler I am sometimes struck by how little other people have experimented but need to remember that stuff I might do in my spare time may have limited appeal for others (I am, after all, a Spurs supporter).
This first video then (4 mins) shows how I might take take some lecture notes (which may be notes from anything of course) and then uses ChatGPT to make sense of them.
This is a guest post from my daughter. It is the outcome of an effort today to kill multiple birds with one stone. First, I have a daughter at home on summer holiday who needs to do more than watch TV or play video games all day and as much as I’d like to go for a long walk or bike ride, I do have work. Second, after overhearing online meeting discussions about plans for new campus spaces the other day she offered some really interesting insights I wanted to capture. Third, in my efforts to find genuine productivity affordances for generative AI I wanted to test the MS Teams transcription function and combine that with ChatGPT text analysis and re-write. So, below is the result. We chatted via Teams for about 15 minutes. I uploaded the transcript to ChatGPT with this prompt: Isolate all comments from ‘Inaya Compton’ then using her arguments and ideas compose a 500 word blog post written in the first person. We then edited the output together and deleted ChatGPT’s typical formulaic conclusion. We decorated the post with a couple of Midjourney images using her prompt: top down view of a classroom with tables in groups of four with two chairs on each side and a board at 90 degrees to the chairs along one wall. Finally, we used the inbuilt AI assistant to review grammar and do a final edit (Paragraph 5 was reworked to be ‘serious’ in tone).
Screenshot from the editor window of this post showing toolbar with three star icon for ‘AI assistant’ highlighted
———————–
Hi, I’m Inaya Compton. As a student, I am enthusiastic and eager to learn, valuing both the intellectual richness of the classroom and the importance of conversation in broadening perspectives. Here, I would like to share some of my insights on the most conducive learning environments, based on my experiences as a learner.
Let’s begin with the significance of the learning space itself. The size of the classroom matters a lot to me. If it’s too large, it creates a sense of separation, while a small space can feel too restrictive. It’s essential to strike the right balance between creating a sense of unity and allowing for individual freedom.
Top down view of classroom generated in Midjourney showing 5 desks each with 6 or 8 chairs around it.
When it comes to teaching small groups, say up to about 10 students, a primary challenge is student engagement and cooperation. The smaller the group, the lesser the interaction dynamics. There’s also a risk of students not learning how to engage with a wider range of people, which is a crucial life skill.
Teachers often struggle with how to address students who don’t readily answer questions. My take on this is that students should have their space. Instead of forcing participation, provide them with an option to indicate their comfort levels with a topic before the session. By giving them an overview of questions and letting them hear other students’ responses, they can still participate indirectly and learn in their way.
When it comes to managing large classes, such as university lectures, adapting the pace to accommodate the diverse learning needs of students can be quite challenging. Within a single class, there exists a wide spectrum of learning speeds, ranging from quick learners to slow learners, with many falling somewhere in between. The task at hand is to ensure that no student is left behind while also avoiding boredom among the more advanced students. Achieving this delicate balance necessitates educators who possess the versatility and adaptability required to effectively cater to the varied learning speeds within the classroom.
Now, let’s talk about the physical arrangement of the classroom. I believe the teacher should remain at the front as it gives a comprehensive range to the class. As for student seating, desks should not be arranged in rows, but instead, every student should be in groups but also able to see the teacher or, more importantly the board or the learning materials. The table configuration in my primary school worked well, with two students on each side of the table, thus avoiding the need for turning around completely.
Top-down view of a classroom, showing a random range of desks of different sizes with varying numbers of chairs and a central paper-covered desk.
It’s worth considering multiple screens in a medium-sized room. Having more than one screen would benefit those seated at difficult angles. And yes, the desks should be movable for a fresh feel every once in a while. It gives a sense of newness, fostering a comfortable learning environment.
Speaking of comfort, this includes the right lighting, temperature, and air flow. A classroom that’s too bright or too dark can be distracting, as can one that’s too cold or too hot. Here in England, for instance, heaters should be used only in winter – don’t keep them on as soon as February hits! Air flow is crucial too, to prevent the room from becoming stuffy.
Finally, seating arrangements should be thought through. My suggestion is to allow students some agency in deciding their seating, depending on their specific needs, such as those with glasses or those who prefer more or less involvement in class. Students with learning difficulties or disabilities should be seated closer to the board and have access to an assistant teacher for personalized help, thereby ensuring that everyone gets the attention they need.
——————-
The various stages we went through to produce this make me think that this will likely become increasingly routine and , although I was keen to really push the generative AI contribuiton, I can already see how the boundaries in some writing between human and AI authorship will be increasingly blurred, hard to track and profoundly complex in providence.
I am sitting in an ALT ELESIG webinar and just saw Dominik Lukes from Oxford University suggest one way students, in all their glorious neurodivergence, could exploit potentials of Generative AI is to reorganise journal content into a different format to aid processing and comprehension. So, no time like the present, I thought I’d try it myself. The idea is to copy/ paste the original text into a tool like ChatGPT then using a prompt like this:
Identify all authors in this text, put them in table and add content to each cell based on the following headings: authors; title of text; key points raised in this article [then pasted in the text from my article]
this is the output:
Authors
Title of Text
Key Points Raised in This Article
Compton (2023)
N/A
Critiques the traditional examination system, citing issues of scalability and pedagogical incompatibility. The pandemic has prompted the consideration of alternative assessment methods, however there’s a renewed call for traditional exams since campuses have reopened.
Williams and Wong (2009)
The efficacy of final examinations: A comparative study of closed‐book, invigilated exams and open‐book, open‐web exams
Closed book exams are deemed outdated in the context of a knowledge economy and counterproductive to constructive learning theories that promote deep learning. Open book exams, which reduce pressure on memorization, provide less opportunities and temptations for academic dishonesty.
Lawrence & Day (2021)
How do we navigate the brave new world of online exams?
Argues that tradition and convenience often inform the continued use of traditional exams. These exams could lead to reduced marking time when feedback is not required.
Durning et al. (2016)
Comparing open-book and closed-book examinations: a systematic review
Supports that traditional exams promote rigor, particularly in medical education. Students tend to study more assiduously for these exams, which could lead to better outcomes for patients.
Johanns et al. (2017)
A systematic review comparing open-book and closed-book examinations: Evaluating effects on development of critical thinking skills
Points out that misunderstandings about open book exams and overconfidence can lead students to waste time searching for answers. Well-designed open book exams can be seen as more authentic, applicable to real-world scenarios, and more learner-directed.
Downes (2017)
University scandal, reputation and governance
Argues that closed-book, in-person exams can reduce cheating compared to other assessment types. Also notes that online exams increase the likelihood of cheating and confusion about what is permissible.
D’Souza and Siegfeldt (2017)
A conceptual framework for detecting cheating in online and take‐home exams
Aligns with Downes, arguing that in-person exams are less susceptible to cheating.
Ottaway et al. (2017)
Cheating after the test: who does it and how often?
Claims that high-achieving students are more likely to cheat on exams.
Simonite (2010)
The impact of coursework on degree classifications and the performance of individual students
Suggests that exams favour high-achieving students, however, the combination of coursework and exams tend to give better results in terms of final degree outcomes.
Williams (2006)
The place of the closed book, invigilated final examination in a knowledge economy
Contests the belief that closed book exams are less prone to cheating.
Chirumamilla et al. (2020)
Cheating in e-exams and paper exams: the perceptions of engineering students and teachers in Norway
Points out that online exams are perceived to be more reliable and efficient, but both staff and students perceive more opportunities for cheating in online modes.
Bengtsson (2019)
Take-home exams in higher education: a systematic review
Argues that traditional exams focus on recall and memorization rather than higher order/critical thinking. Such exams also put unnatural pressure on students, impacting their performance.
Clemmer et al. (2018)
Will that be on the exam?-Student perceptions of memorization and success in engineering
Points out that students using memorization techniques may perceive exams as unfair if questions do not reflect content they have practiced.
Rich (2011)
An experimental study of differences in study habits and long-term retention rates between open-book and closed-book exams in a water resources engineering class
Suggests that open-book exams encourage students to adopt more effective study habits and demonstrate better long-term retention of course material.
For me, this would have been a complete game chnager when I was a student, espceilly when conducting literature reviews.
Select the Content: First, you need to choose the article or content you want to summarise. The content could be a journal article, a blog post, a news report, or any written material.
Copy the Content: Once you have the content, copy the entire text or the relevant portions you want summarised. This is your raw material for the summarisation task.
Navigate to ChatGPT: Open a new browser window and navigate to the ChatGPT interface. This could be a website or an application where you can interact with the ChatGPT model.
Paste the Content: Paste the copied content into the chatbox of ChatGPT. You need to be careful not to exceed the maximum character limit allowed by the tool.
Create a Prompt: After pasting the content, create a prompt for the model to structure the summary. Here’s an example prompt: “Identify all authors in this text, put them in a table and add content to each cell based on the following headings: authors; title of text; key points raised in this article.”
Run the Model: After inputting the prompt, press enter or click the appropriate button to execute the command. The model will process the input and produce an output based on your prompt.
Review the Output: Review the generated summary and make sure all the key points from the original content are captured accurately. If necessary, refine your prompt and rerun the model.
Copy the Output: If you’re satisfied with the output, copy it for use in your desired application.
Refine and Iterate: Keep in mind that AI models like ChatGPT may require a few iterations to get the desired output. Don’t hesitate to refine your prompts and iterate the process.
This method will allow you to create a structured summary of any written content. Note that the AI will follow your prompts, so be as specific as you can to get the best results.
This open crowdsourced collection presents a rich tapestry of our collective thinking in the first months of 2023 stitching together potential alternative uses and applications of Artificial Intelligence (AI) that could make a difference and create new learning, development, teaching and assessment opportunities. Experimentation is at the heart of learning, teaching and scholarship. Being open to diverse ideas will help us make novel connections that can lead to new discoveries and insights to make a positive contribution to our world. Ideas shared may be in its embryonic stage, but worth exploring further through active and creative inquiry.
With the publication, we would like to illuminate the importance of responsible, critical and ethical use of AI in education settings and more generally.
We are grateful for all 101 contributions from 19 countries: Australia, Canada, China, Egypt, Germany, Greece, India, Israel, Italy, Ireland, Jordan, Liberia, Mexico, South Africa, Spain, Thailand…
Teachers who care, who serve their students, are usually at odds with the environments wherein we teach (hooks, 2013)
It is very easy in a modern university to get absorbed with systems, processes, data and results that often de-centre the individuals that work and study within these systems. We hear increasingly of the troubling consequences of student wellbeing issues and of staff burnout, and the pandemic has exacerbated many of the tensions and issues consequent of highly-pressurised ways of working and being that are common in higher education. A pedagogy of care deliberately pushes against these pressurised phenomena. It centres individuals by starting with respect, trust, inclusion and relationship-building as precursors to dialogue and affective development as well as academic development.
even for the majority who do “care” in the virtue sense—that is, they profess to care and work hard at their teaching—there are many who do not adopt the relational sense of caring. (Noddings, 2005)
As a prompt for discussion and as a starting point to help us all (as educators working in HE) interrogate our own current practices, we offer the following ‘pedagogy of care (hu)manifesto’ which draws on core concepts, principles and ideas found in the works cited below. We invite colleagues to consider their own (and their peers’) practices in light of each of these statements, to identify tensions, challenges, objections and potential pitfalls as well as opportunities, examples and affordances suggested by each of the commitments.
By embracing a pedagogy of care, we endeavour to:
1.
Humanise things! Understand the value of connecting at a human level and modelling caring
2.
Challenge conventions of hierarchy and authority
3.
Challenge the narratives and norms of rigour and educational ‘suffering’
4.
Normalise learning through mistakes
5.
Recognise that positive relationships demand trust: Being ‘nice’ does not mean being indirect or dishonest
6.
Appreciate that dialogue is essential to showing care (and listening is at least half of this!)
7.
Accept that humility and normalising vulnerability show strength not weakness
8.
Show and tell students that you care- DO smile before winter break!
9.
Employ flexibility, openness and welcome with office hours
10.
Above all: acknowledge where each student is at and don’t enforce behaviours or punish recalcitrance
In the case of wellbeing interventions in higher education, lesson- learning, sharing good practice and building networks around ideas and interventions are all important, but it is also critical to understand factors that shape HE organisations’ abilities to successfully take this knowledge forward and address wellbeing problems. (Watson & Turnpenny, 2022)
hooks, b. (1994) Teaching to transgress. Oxon: Routledge
hooks, b. (2013) Teaching community: A pedagogy of hope. Routledge.
Hughes, G, Upsher, R, Nobili, A, Kirkman, A, Wilson, C, Bowers- Brown, T, Foster, J, Bradley, S and Byrom, N (2022)Education for Mental Health. Advance HE.
Not a week goes by when I am not bowled over by the brillliance of students (and colleagues too of course). Recently it has been Kaz and Navi whose articulate ruminations on generative AI are really thought-provoking. Last week I was blown away with the work of Olivia Moir, a student at the UCL Institute for Women’s Health, who was keen to start a podcast series. I’d previously shown the workings of the Arena podcasting studio but I wasn’t expecting such a swift application of those developing skills and a very professional first in a series podcast. Olivia writes more below:
This past week I launched my new podcast, It All Starts Here, for the UCL Institute for Women’s Health. The podcast aims to provide a platform that brings awareness to various critical topics in the fields of reproductive science and women’s health – topics that are often not discussed enough. We cover a range of topics and focus on creating content that is understandable to individuals from all levels of education and backgrounds, with no science experience required.
This is something I’ve been working on since last year, and have put my entire heart into, so am very excited to finally share it with everyone!
Admittedly, in the initial stages of figuring out the logistics of this podcast, my biggest concerns were around my lack of skills when it comes to technology. Thankfully, these anxieties were quickly put to rest following a couple of training sessions with Martin, who showed how intuitive and easy to use podcasting equipment can be.
This has been a wonderful journey so far, and I have already learned a lot. Really looking forward to continuing with this and seeing where it goes!
