Also check out AI@UNE Part 2 | Part 3 | Part 4
Generative AI is taking the world by storm and its impact on higher education will be profound. At UNE, staff are exploring innovative ways to embrace its capabilities while also considering its misuse and potential impact on academic integrity.
Lecturer and Academic Integrity Officer with the UNE Business School, Aaron Driver, has been delivering webinars to UNE staff on generative AI. In the first instalment in our new AI series, we asked him to bring us up to speed with this fast-evolving technology.
NB: Since this series was launched, Aaron Driver has been appointed as AI Integration Lead, Faculty of Science, Agriculture, Business and Law.
What is generative Artificial Intelligence (AI)?
Generative AI, of which ChatGPT is currently the highest-profile example, generates new content based on patterns it learns from a large amount of data. In the case of Large Language Models (LLMs) like ChatGPT, it’s akin to an advanced predictive text system. It learns language by being trained on vast amounts of text data from the internet (like a child reading lots of books to learn language), predicting the next word in a sentence, and gradually improving over time.
AI learns from text that comes from a wide variety of sources on the internet, including from universities. But AI doesn’t know specifics about which documents or authors it learned from. Once trained, the AI creates new sentences by taking an input (like a prompt from a user), ‘thinking’ about what it has learned, and writing a response. Importantly, it doesn’t truly understand the text it generates. It’s essentially a complex mimicry tool that crafts human-like responses based on patterns it has observed.
It doesn’t truly understand the text it generates. It’s essentially a complex mimicry tool.
The technology is moving at break-neck speed (ChatGPT is the fastest-growing product in history), and capabilities and use cases are multiplying and diversifying. The technology doesn’t yet appear to have reached a terminal velocity, i.e., the speed of change is continuing to increase.
Aaron Driver.
What broad impact will generative AI have on society?
What was the impact of fire as a technology? It allowed us to consume more calories (by cooking our food) and increase our REM sleep (less predation at night), which resulted in massive growth in the size of our brains, and everything that followed from that. Arguably it’s hyperbolic to compare AI to fire, but maybe not, and the point is that the average human in the Palaeolithic would have struggled to predict fire’s long-term impact.
But to hazard a short-term prediction … Accenture Research, for example, has predicted that LLMs can impact 40% of working hours across industries. This is based on language tasks accounting for 62% of total worked time in the US. Of the overall share of language tasks, Accenture estimates 65% have a high potential to be automated or augmented by LLMs.
Tl;dr: A big impact!
What are the specific implications for universities?
Generative AI represents the biggest shock to the higher education system since the internet, and I think it poses a potentially bigger threat to the way we educate students. The written word – text – is now a commodity. And we know our students are using it – many of them are paying for the plus version of ChatGPT – for ideation, research, rewriting and, unfortunately in some cases, to cheat.
Although there is no shortage of threats, we are also very much focused on the opportunities.
Many traditional assessment methods can now be performed by AI, requiring a fundamental re-evaluation of assessment design. Current detection tools have their limitations, so UNE Business School is creating and implementing a new assessment design framework. This framework aims to devise assessments that are either exclusive of AI (that is, the AI can’t help students unfairly) or are AI-inclusive (where students are encouraged to work with the AI to complete the assessment).
Although there is no shortage of threats, we are also very much focused on the opportunities. Generative AI can help us do more, better, and faster. And let’s face it, our students will be working with this tech once they graduate. So, we’re rolling up our sleeves and getting them ready for that future.
