ChatGPT3 was launched in November 2022 and has sparked a huge amount of investigation and conversation about its potential role in education. From our vantage point of April 2023, what might the impacts on pedagogy be?
Five key elements seem to be emerging:
1. Let’s start with the wild hope that maybe this time we’ll do something about assessment practices.
During the pandemic, AI proctoring called into question our substantial addiction to invigilated exams. ChatGPT is now calling into question our penchant for formulaic essay writing. Already, there are decades of research supporting change here, which on the whole have been ignored.
Authentic assessment, reflective learning, ungrading, developing evaluative judgement: Every one of these practices would support the development of lifelong learning skills. But how many educators have that in their learning and teaching strategy?
2. Next we might need to reconsider what we mean by “academic integrity.”
When knowledge can be co-constructed with tools like ChatGPT, what does that mean for “authorship?” Did you really write that course syllabus? Did your student “collude” with ChatGPT to write that essay?
An interesting conundrum recently posed on Twitter asks us to think deeply about what “original” work means, and the relationship between assessments and learning outcomes.
“Student writes essay in first language that’s not English in a science subject. Uses Google translate to put it into English. Uses AI tool like ChatGPT to smooth it. Detection tool + ‘gut feeling’ says this isn’t this student’s work. Is it? Is it cheating?” (https://twitter.com/mart_compton/status/1629412722373799938)
We know that good scholarly practice avoids plagiarism, but where we choose to incorporate a tool like ChatGPT we are advocating the use of something that has effectively plagiarized on an unfathomable scale. We are also selling ourselves and our students’ labour as yet another source of the human “ghost work” that exists behind much AI.
3. We might also need to consider our ideas about who the teacher is.
AI forces us to go beyond the technology as tool metaphor and think about complex entanglements of humans and technology. Where does the human teacher end and the digital teacher begin? What are the interplays and relationships between them?
“While we work with the present (generative) incarnation of AI technologies, we see a need to identify effective learning and teaching practices that will harness the weaknesses of generative AI technologies as opportunities for promoting higher-order learning (e.g., analyze and scrutinize outputs produced by ChatGPT).” (Gasevic et al. Empowering learners for the age of artificial intelligence)
If a student spends time asking ChatGPT questions, evaluates the answers and in doing so learns something new, can we say that ChatGPT is fulfilling some of the role of a teacher?
4. We can take opportunities to build digital literacies.
Students live and work in increasingly technologically mediated worlds. To function effectively and to flourish, they need more than just knowing how to use the Microsoft Office suite. Being able to confidently evaluate technologies and make judgements about their trust and reliability will be crucial skills.
Can we spend time talking about how tools like ChatGPT work? Can we ask: Where the data comes from? How do these tools generate text patterns and not meaning? What biases do they contain?
“GPT and other large language models are aesthetic instruments rather than epistemological ones. Imagine a weird, unholy synthesizer whose buttons sample textual information, style and semantics. Such a thing is compelling not because it offers answers in the form of text, but because it makes it possible to play text — all the text, almost — like an instrument.” (Bogost, ChatGPT is dumber than you think)
5. Finally, we can consider our fundamental roles as education institutions.
As hinted at in all the points above, the use of a tool like ChatGPT is not unproblematic. It challenges us to consider what we value and how we work. In using these technologies, we have to consider how they came into being, and how they are being positioned in terms of futures. Beyond working out whether to adapt our learning and teaching to take advantage of them, we have a fundamental duty as education institutions to resist moral panic and inaction, and instead grapple with and make sense of them. This may be the biggest contribution of all to learning that we can make.
AI technologies are not new, and many more will continue to emerge. Indeed, ChatGPT4 is already on the horizon, and there is a whole slew of advice out there on other AI tools that can be used in academic work. The change we need to make to adapt to them is to be continually ready to reflect on what learning is, how it happens and how we work as education institutions. The answers will not be simple, the work required to adapt will not be easy or cheap, but the potential is enormous if we choose to embrace the change.
Fundamentally, new AI technologies absolutely have the ability to transform education, whether we choose to adopt them or not.