ChatGPT Forum - Massey University

 Notes from a zoom presentation hosted by Massey University on ChatGPT - a cross universities' forum.

Professor Giselle Byrnes begins with an overview of the technology and introduced the speakers. Each speaks for 10 minutes and a Q & A session follows.

Dr, Mark Nichols (Open Polytechnic / Te Pūkenga) begins by starting that AI has no ethics, morality, conscience. Therefore, it should not be anthromorphised. AI can be used for good or for evil but has great potential for education. AI draws on two negative traits - one is that we think we know more than we know and the other is that we are programmed to take shortcuts. Therefore, AI needs to introduced carefully with understanding of its potentially and limitations. It is important to realise that although the outputs of ChatGPT look impressive, they are not always correct. It lies and makes up references. 

Undertook an overview of the Polytechnic's learning design process and how AI can be used to augment their work. User education must be undertaken to address deception and ignorance - the enemies of scholarship. AI / not-AI boundaries are now blurred, banning AI is naive. 

Dr. Karaitiana Tairu continues by bring a Māori lens to the discussion. Overviewed the constitutional and legal considerations on data sovereignty. the use of Māori data can be considered as drawing on Māori taonga. Contends that there is a re-colonisation of Mātauranga Māori as it is not part of the ethics, etc. of the algorithms behind - for example - ChatGPT. At the moment, much of Mātaurange is not digitised, draws on Māori data and ignores Māori data soverignty. ChatGPT is creating false Māori histories and produces pepeha / karakia with mistakes!! Intellectual property rights and Te Tiriti are not followed. However, ChatGPT is relatively good at te reo!! could be useful and contribute to learners of Māori. 

Lee Hicken (microsoft) acknowledges the sense of fear and awe with regards to how AI has developed and advanced. Used ChatGPT to synthesis some thoughts. Referred to Atthur C. Clarkes 3 laws to describe what is now occurring. We cannot block or go back. How can we move forward?? AI does not know anything, but has a strong 'grasp' of how language works and therefore uses linguistic patterns to work out 'answers'. Generative AI is now a blank piece of paper, when we ask it questions, we train it. It learns our biases etc. It is therefore our responsibility to ensure that how we use ChatGPT is done in a carefully thought out manner. We are its teacher!! and it will reflect our thoughts. It is important to teach our students how to work and live with AI. In particular to ensure critical thinking is a key to how we use AI.

Then, Dr. Colin Bjork from Massey University presented on AI and cheating. Rather than stop students from cheating, it is important to look into why students are shaped by inequality / access due to issues of unresolved inequities etc. Experiences that shape our perspectives on the world - language, gender, age, ethnicity, social economic status etc. etc. also colour how we may interact with AI, leverage off its uses, and understand and think through the implication of AI. Banning AI in education will likely hurt those who may benefit from its use (those with disabilities etc.). Worries that the language used by AI becomes the status quo and that access/lack of access exacerbates the inequalities that already exists. 

Graeme Smith from Ako Aotearoa. The future is here but it is unevenly distributed. Summarised and discussed the various items introduced by the other speakers. What can teachers working with a class do? ChatGPT will not likely replace a teacher. Working with ChatGPT can be used to augment our work as teacher. AI will only accelerate from this point on. Learn early and be familiar with the tool. Cool stuff can be build using Chat GPT. We need to work in conjunction with AI. When we can 'clone' ourselves with our own customised AI, we may offload the 'boring' stuff and indulge in the creative and interesting things.

The webinar closed with encouragement to continue the conversation, opportunities for research and dialogue. Another channel opened further discussion.