by Bethany Witham and Rachel Bancroft
Given our fascination with digital pedagogies, online learning and teaching and the many ways education is evolving and changing, it is probably no surprise that the members of the LTSU are voracious consumers of journal articles, blog posts, and Toots and Tweets a-plenty on all things Learning Technology. As such, we’ve decided to start sharing our thoughts on some of the things we read in regular #LTSUReads posts on this blog. As well as giving a sense of our interests, and hopefully piquing yours, it’s also a great opportunity for us to share some of the things we’ve found out about new and emerging technologies, or current events that are likely to have an impact on online learning and teaching. That’s why this post will be a little longer than usual – we wanted to take this opportunity to share some of the things we’ve learned about the new Artificial Intelligence Chatbot, ChatGPT.
What is ChatGPT?
ChatGPT, according to the bot itself, is “a state-of-the-art language generation model that can generate text based on a given prompt or context.” Given its seemingly human responses to a wide variety of questions and prompts, it has become a massive subject of conversation recently involving many groups in higher education. Observations we’ve read so far are generally emerging thoughts based on a relatively short period of access to the technology, so it’s too early for definitive facts about what ChatGPT can and cannot do. Given that it is being trained on the prompts given to it by the public, its precise functionality could be continually shifting anyway. To find out more about how ChatGPT was trained, a good place to start is Michael Webb’s article GPT-3 and Plausible Untruths.
We’re participating in the conversation and doing lots of reading to try to understand what thoughts and themes are emerging; in particular, we’re conscious of the thoughts of those advising caution about using the tool, so this list of links should not be seen as advice on what to try! Instead, we have tried to share an introduction to the conversation; we’ll be exploring what it means for us in our School in a ‘Time to talk’ session for NTU Arts and Humanities colleagues.
Where to find out more…
Jisc: How to teach your cat to code on the Jisc blog by Michael Webb
“The way students find information and answer assessment questions is already changing, and education will need to adapt in response. We should see these new developments as an exciting opportunity to really rethink the future of the assessment process.“
Writing for Jisc, Webb provides a brief introduction to the technology and a sense of what it can do, and where its limitations may lie. Rather than offering any conclusions, Webb offers a place to start – learning and understanding more about the potential of ChatGPT can help us to make decisions about how to make the most of it.
GPT-3 and Plausible Untruths for the Jisc National Centre for AI by Michael Webb
A follow-up to Webb’s Jisc article linked above, this explains in more detail how ChatGPT formulates responses to the prompts it is given and gives examples of where the AI has created ‘plausible untruths’.
ChatGPT and the rise of AI writers: how should higher education respond? On THE Campus by Nancy Gleason
“Assessing only a completed product is no longer viable. Assessment needs to shift to process. This has always been the case, but ChatGPT is forcing the issue. Scaffolding in the skills and competencies associated with writing, producing and creating is the way forward.”
Advising against kneejerk reactions to the advent of ChatGPT, in this article Gleason gives an overview of the technology and starts to consider the implications for assessment design.
Robot-generated submissions on Assessment for Higher Education, a blog post by Rachel Forsyth
“New thinking isn’t needed, it’s just a prompt to do the thinking which should have already happened. We should now review our assignment tasks* to ensure students are using all the tools available, including AI, to demonstrate their high level thinking.”
Forsyth offers practical tips about how to design and update assessments to mitigate the risks of students using AI-generated responses, providing suggestions that also have the potential to be more inclusive of students with non-traditional routes into university, or those with specific learning differences.
Who’s afraid of ChatGPT on Teachers Going Gradeless by Martin Compton
“…above all, finding ways to de-center the grade and re-energize student interest in the process not the product of their education should be driving us all right now.”
Compton offers a personal account of how he believes new tools should be incorporated into existing learning and teaching approaches, helpfully contextualising the current worries about ChatGPT by likening it to other ‘disruptive’ technologies that have caused similar concerns – such as the calculator, or spelling and grammar checkers. Overall, Compton says “My message then, as now, was ‘Don’t panic: let’s talk this through.’”
ChatGPT Discussion, a video with Dave Cormier and Tim Fawns
Rather than a close focus on an immediate reaction to the specific disruption tied to the advent of ChatGPT, Fawns and Cormier discuss wider issues related to the continuing impact of technology on education, and draw attention to the purpose of assessment within education, and how it serves the learner and educator both. This video is an invitation to pause and reflect on our practice in learning and teaching, rather than a push towards a reaction to what is essentially just one new change in an ever-moving field.
I Taught ChatGPT to Invent a Language on Maximum Effort, Minimum Reward, a blog post by Dylan Black
“In which ChatGPT and I invent a fictional language spoken by slime-people.”
While perhaps less related to the impact on education, this is an interesting read for all our linguists and language experts!
Come and talk to us!
As always, we are available to discuss this further with our colleagues in Arts and Humanities. Please email us at AAH.LTSU@ntu.ac.uk.
