Tertiary Education Union (TEU) AI conference

 The TEU- Te Hautu Kahurangi -  convened a conference on AI, with panels discussing a range of topics related to AI, Union support, and teaching and learning. Due to other work commitments, I am only able to attend two of the sessions and be part of the panel on Mana Mātauranga (power of knowledge). 

There are sessions on Mana mahi - keeping decent work at the centre as AI reshapes tertiary education; Mana Taurite - exploring how AI can support equity and inclusion across tertiary education and Mana Taiao - ensuring AI in tertiary education aligns with climate action and environmental justice.

Notes taken from the sessions I was able to get to are below:

Seesion on Mana mahi - keeping decent work at the centre as AI reshapes tertiary education.

Dr Hansi Gunasinghe (Southern Institute of Technology). - Ai transforming education from administration to research. How can we balance innovation with dignity of work? Need to discuss the understanding of Gen AI and the human role. Went through principles of Gen AI - what it is, main types, applications. Shared the examples, opportunities and risks with Gen AI. AI can support teaching - tools help create lesson plans, quizzes, tutorials, and accessibility resources. Automation should support educators, not replace them. Shared research underway with students in China studying how to design a mobile application prototype using FIGMA. AI was used to create marking criteria with human evaluation from tutors. Automated evaluation undertaken using ChatGPT, Gemini, Copilot and compare these with human evaluations. Conclusions will derive time used, accuracy and the process. the next step is to integrate FIGMA prototype extraction tool and video creation. Ai is fast, scaloble, data drive. Humans are empathetic, ethical and contextual. A balance defines the dignity of work. Ran through current institutional policies on AI use for students, the policies for staff are still under development. Therefore, AI must be a support not as a 'supervisor'; ensure human oversight and co-design with educators and learners. Adopt transparent, open and ethical systems. Dignity in work means that humans lead. This needs to preserve profesisonal judgement and relational teaching. include Ai literacy and conultation in policy. Respect Te Ao Māori principles and protect workload fairness and AI should strengthen, not weaken, our human values. 

Dr. Leon Salter (University of Auckland) - AI working group at UoA - summarised UoA approach, with his personal viewpoints. VC's forum and new action plan shows the university to be techno - optimistic. The action plan was prepared by ' the AI education advisory group' which did not consult with staff, unions and chaired by the Director of Learning and Teaching. The document mentions risks and guiding principles at the start but never mentioned again in the rest of the 10 pages! In general, to exhort staff to integrate AI into teaching and learning. Offers 'carrots' to encourage AI integration with underlying disparaging of reluctant staff as 'dino-professors'. A AI working group formed which is open to all TEU members to  provide critical feedback on the university's direction and be a support group. Summarised work of Dan McQuillan on 'resisting AI - an anti-fascist approach to AI'. Shared the findings from a TEU survey of university staff to find out about their perspectives on university policy and communications on AI and usage of AI in workplace. Overall sentiment of discontent - generally dissatisfaction with leaderships communication and policy. Open to sharing perspectives from others running similar groups, and keen to connect. 

Session on Mana Tiriti.- navigating the relationship between AI, Te Tiriti o Waitangi, and the future of tertiary education

On the panel are Brendon Shaw from Papatoeotoe North School, Dr. Kevin Shedlock from Victoria University Wellington, Olivier Jutel (University of Otago) and Warwick Tie (Massey University).

Brendon presented on 'He whānaungatanga Tīmatanga: AI and Te Tiriti o Waitangi in our schools'. He teaches at a primary school but is also a PhD studenta at University of Waikato. Went through the ways principles of Te Tiriti are enacted in schools with respect to AI. With Partnerships - Māori need to be involved in the co-design of the AI lifecycle from data to deployment. True partnership means Māori are present in the creation of AI systems, not just assessing outputs. Māori-led initiatives like Te Hiku Media exemplify bicultural AI governance. Participation means removing the barriers to access the technology. Work needs to be undertaken on the digital divide and equitable solutions. The CoVID pandemic showed how Māori are on the wrong side of both the digital divide and equity and these need to be addressed. Protection includes being cognisant of safe guarding Taonga and identity, as expects of data sovereignty need to be culturally relevant. Data still biased and created images based on these.

Kevin's topic was 'navigating the relationship between AI, Te Tiriti o Waitangi, and the future of tertiary education. Covered Te Tiriti when creating digital AI artifacts; indigenous Māori knowledge that resides in these; and the digital artifact. Ai in tertiary education can be envisaged as being a conduit where both indigenous and western research paradigms are able to reside in search of new knowledge within the Tiriti. This is challenge as there are not enough people on the ground to support the indigenous side. Challenges also include digital inequalities, unequal access to heritage, power imbalances in defining narratives, data ownership and digital ethics, lack of adequate infrastructure and environment pressures and sustainable issues. So what is good vs bad Māori data?? Difficult questions requiring deep understanding and immersion, not just to tick the box. Indigenous Māori knowledge in artifacts needs to go beyond the surface. There are no protocols for where Māori knowledge resides and therefore its expression in data is different, its needs to have the correct framing, be build on relationships and thrive through engagement, not as a piece of digital artifact. Check with work of Shedlock,R. and Hudson, P (2022) - use of Mātauranga clearly organised Advocates for reciprocal and respectful AI. Requires good understanding of the problem; trusted relationships and mutual agreement to reach consent. 

Oliver  - we are not able to build Mātuaranga Māori onto the existing western techno-centric AI model. Spoke on 'AGI and the PE of AI'. Ran briefly through the ideological history of AI - George Boole, Marvin Minsky and ELIZA as the foundation of the western technocratic understanding and roots of AI. Shifted to the current day, with perspectives on AI from Musk, Altman. Book by David Noble 'The religion of Technology recommended. Connects religion and AI marketing. Detailed how AI is overblown with not much making sense with regard to AI possibilities for making money. AI must not be thought of in 'god-like' terms but challenged to make contributions to the wider society. 

Session on Mana Mātauranga - harnessing AI to strengthen teaching, learning and research, and uphold public tertiary education.

With me on this panel are Dr. Shahper Richter (University of Auckland), Dr. Warwick Tie (Massey University) and Traci Meek-Reid (Southern Institute of Technology) who facilitates the session.

I begin the session with an overview of the AI@Ara projects. Summarising their overall objectives, pedagogical underpinnings and implications on the work of tertiary educators. Details of these the two main projects are now published in the book - AI in vocational education and training; and the Ako Aotearoa report on 'AI to support foundation/bridging learners'. It is important to undertake good learning design so that AI does not replace the 'learning activities' required for learning and teachers role is well defined. 

Warwick presents on "upping our game" - shared the themes from his course on AI. AI's relation to language through speech acts, discursive formations and the discursive. Ai in the discursive can be envisaged as AI as the 'public unconscious' and suffers no anxiety! AI does not use language, but redefines how we understand, teach and use it. Ai is a player in the linguistic landscape. In his course, how can we do more with words (upping our game).  Therefore to view AI as static (postivism); AI in movement (dialectics); Ai as a discursive construct (discourse analysis). Showed matrix on how each of these affected by AI across various social categories. Discussed the assignment - asked students to use the Gen AI to write a 1000 word essay on what needs to be asked and then analyse what the bot has provided (1500 words). Is the AI approaching from the positivist, dialectical or discursive approach?? AI agents can only deal with what it can see, but not it cannot. It can discuss the doughnut but not the hole in the middle :) 

Shahper continues with discussion on Warwick's assessment approach. The session moves into a discussion on AI in teaching and learning and the various ways assessments can be shifted to include AI or be used to learn the limitations of AI. The conversation moved to what AI to use and when it can be used. Assessment standards are not much higher as the assumption will be that students have access to AI and will use them. Digital equity discussed as the challenges are different across the sectors. 

Dr Julie Douglas TEU Te Tumu Whakarae (National Co-President - Tiriti) facilitates the last session for each of the facilitators through the day, summarises the sessions they have chaired. 

All in, good discussion, with many perspectives from across education sectors, on the opportunities, challenges, implications and promises of AI in education. The advantage of the discussions is the focus on being critical educators, leading the integration of AI into education. We must actively contribute to the policies and conversations around AI, providing a voice for educators and learners. 

The conference closed with the reading of the TEU karakia.

Overview of using various AIs - Professor Ethan Mollick

With so many AI choices out there, it is difficult to select one which will effectively support the task at hand. Professor Ethan Mollick who was a very early adopter of Gen AI in the higher education sector, provides an overview of AI as they stand near the end of 2025. 

The blog compares Gemini, Claude and ChatGPT and their uses in an academic context. The various categories of chat, agent and wizard models are compared. Inputs of text, voice and images along with outcome in images, video, code and documents are also discussed and evaluated.

As always, there is not one Gen AI platform which will 'do it all' or do all well. Therefore, it is still important to match the objective / end goal, with the Gen AI. As Gen AI develops rapidly, it is also important to keep up with the play. For example, we have been encouraged to use CoPilot at my institute. To start with, CoPilot did not really compete well with the other mainstays - ChatGPT, Claude and Gemini. However, CoPilot has improved markedly and is now integrated into the many Microsoft tools we use daily. The M365 version of CoPilot provides ability to create agents and also a range of agents which are useful for common everyday tasks. Paying extra includes the research and analyst agents. These two agents are extremely useful not only for research but for general tasks requiring inquiry, comparisons and deeper evaluation. Therefore, it is important to keep up with the play!! 

Aotearoa Tertiary AI network ATAIN - presentation from Dr. Simon McCallum

 Notes taken from presentation from Dr. Simon McCallum, Victorial University Wellington on 'Adapting to AI'.

The presentation is part of a fortnightly series organised through ATAIN which is a SIG of Flexible Learning Association of  NZ (FLANZ).

Simon began with an introduction. He has been teaching game development since 2004 but has also taught AI since 1991. Noted that Gen Ai is everywhere and we use it unintentionally, unconsciously, but also using it consciously and strategically. Productivity benefits depends on training of the AI. Ai agents work together to automate generic tasks.

Across industries, adoption is mixed, some fast, some very slow. The risks include programmers with AI automating other industries and the use of 'software / automation on demand'. 

Revised the two lane approach to assessments. Students are using AI - At UVW 66% admit using it.

Covered the following:

 AI literacy - Core / domain specific - compulsory for all students and staff, understand if and when to use AI and avoid the risk of thoughtless AI use.

Assessments need to move to testing understanding and learning rather then outputs. Test meta cognition, use oral assessments.

All non-invigilated (lane 2) work should be considered as group work, using group work assessment techniques - assess process, influence, delta, learning journey. 

Assessments can embrace AI assistance. AI selects questions, create bespoke questions, suggest oral assessment questions. Human markers determine the grade. Provided details of the process from his context. Use AI to generate questions from student submission for them to complete, to check understanding they have presented in their essay!

Increase the quantity of group/team work so students can create human connections and increase their experiences with working with others. Invrease the amount of work that is groupwork but not the groupmark.

Proposed that NZ universities fund a NZ based server to assure AI sovereignty. A more equitable approach as all students will have access to high quality models instead of having to pay for the upgraded models. 

Also the creation of a position to report to Te Hiwa (leadership group to organise and manage AI across the entire university - teaching, learning, research and professional. 

Summarised agentic AI. Moves AI from being a chatbot to actually being able to 'do things'. AI will sort out a plan, and work through it to meet the prompt objective. Swarm coding can be activated, to check the outputs from each agent. Therefore, AI is not a search engine. It is better to have the AI question us to work out what we want done! 'help me do ----' 

Summarised project on how AI is used in NZ secondary schools. Mixed across schools on strategy, current use, professional / student use and community involvement. Schools welcome clearer policies and guidelines. Challenges are similar to Universities, assessments, professional development, etc.

Shifted to a summary of the extrinsic and intrinsic purposes of learning. If the motivation is just to pass the exam, then AI is an impediment. However, if there is a less constricted 'assessment' e.g. develop a game, AI accelerated capability and leads to extended learning. 

Therefore important to engage learnings to focus on intrinsic motivation, self-reflection and etc. and to hold them to account for what they want to  learn. Teacher as NPC . Encouraged learner negotiated assessments and rubrics, giving them agency. For example, have learners establish the range of marks assigned to various aspects of an assessment. This encourages meta cognition for students to structure their learning, their strengths/ weaknesses etc. 

How do we measure metacognition - confidence is good when it is accurate, under / over-confidence is a problem and AI makes this worse. Important to operationalise and elicit accurate statements from learners.

Invitation to use the group's Discord to continue the conversation and share ideas. 

AI-generated assessments for vocational education and training - webinar

 Here are notes from the webinar on the ConCove Tūhura project AI-generated assessments for VET, 

The report provides the literature scan and details of the process undertaken to identify appropriate AI to undertake the task, and the processes to ensure that the AI- generated assessments would meet moderation requirements (quality assurance) for use for assessing VET standards. 

The work was undertaken by Stuart Martin from George Angus Consulting and Karl Hartley from Epic Learning. Both present in the webinar which begins with an introduction by Katherine Hall (CE for ConCoVE Tūhura) and by Eve Price (project manager at ConCoVE).

In Katherine's introduction, the rationale for the project was shared along with some of the journey taken by the project to break new ground.

Eve Price provided the background of the project. Most projects focus on integrating AI into ako or the prevention of AI for assessment. This project wanted to help support the time consuming 'back room' processes including resource and assessment development.

Karl ran through the approaches to the product. The evaluation/review processes could not really keep up with the speed at which assessments can be developed when it is supported by AI. 

Stuart shared reflections on how the process evolved and the various processes put in place, were reflected on and were then reintroduced into the AI-generation project. Explained how various quality pointers were met to ensure the efficacy of the process.

Eve detailed the need to be specific with what needed to be achieved - assessment, feedback, etc. Selection the correct AI is also important. Prompts are detailed in the project report. Important to evaluate at each step.

The bigger picture with micro-credentials, skills standards and AI-generated assessments all add innovations to the VET ecosystem. Understanding the policies and processes used by WDCs and NZQA need to always be part of the process, so that various quality points are met.

Stuart summarised some of the challenges and how the project worked through these. 

Karl talked on the importance of people in the process when AI is generating the assessments. Firstly, important to understand some of the mechanics of AI - what is under the hood. Secondly, quality assurance must be focused on the concepts, not so much the grammar/spelling etc. Thirdly, need to make sure assessment purpose is clear. 

Next, academic integrity and ethics were discussed. Important to ensure that there is understanding the impact of AI on privacy and data sovereignty (including indigenous perspectives). Important to train the AI to understand tieh Aotearoa context. Claude AI was selected due to its stance on human rights, ethics etc. 

Findings included: assessments did not meet moderation but improved the opportunities for inclusiveness and personalisation of learning. Failing moderation added to the learnings from the project. The items involved too many questions, answers being at too long and at too high a level. 

Eve reiterated the need to 'define what good looks like' to the AI, so that human objectives/ perspectives are taken into account. Important to ensure principles of ethics etc are maintained as it is important to 'keep humans at the centre'.

Karl's learning include AI drawing in novel content through its hallucination. The AI included assessor approaches into its assessment and this caused him to consider the learner information that should be included to provide direction. The U S of A standardised approaches to writing assessments, seemed to permeate the assessments produced by AI. This had to be superseded through careful prompting.

Flexibility to allow for personalisation to industry (example safety unit standard customised to a range of work roles/ disciplines); and learners (for ESOL, neurodiverse learners etc.). 

 Q & A followed 

The webinar was recorded. 

Discussions revolved around practicalities, challenges and solutions.

All in, good sharing that adds to everyone's learning about the roles of AI to support teaching and learning, integration of practice/practical and cultural contexts, the need to be aware of the fish hooks' in using AI, how quickly AI is developing to meet user needs, and the need to continually learn to ensure that the understanding of AI / ethics etc. form the foundation for working with AI. 

AI forum productivity report for New Zealand

This report - AI in Action: Exploring the impact of AI on NZ's productivity, is produced by the AI Forum NZ in partnership with Victoria University Wellington and PR powered by heft.

It is the third biannual report and collates an overview of the impact AI is having on productivity across NZ. Since the first report in 2023, there has been growth in the use of AI with accompanying effects on work, the workforce and contributions to the economy. 

The 3 page executive summary provides the main points. Key findings are then extended and discusses followed on by case studies.

In the businesses surveyed, 91% reported productivity gains, 50% view AI contributes to cost savings with 77% saying that there have been actual cost savings. However, less than 25% had savings over $50,000. Therefore consistent productivity gains.

Workforce impacts include increased job losses which reflect the country being in recession; 55% reported that new roles have been created; cost of setting up AI have reduced, strategic policy investments have been attained; operationally, AI cost less. 

AI's introduction requires the building of trust across the workforce with AI literacy being a key and the need to ensure that there is inclusive engagement for all.

Overall, data that reports on growing adoption, settling in of organisations into understanding how AI can be leveraged to increase efficiencies, and acceptance of AI as inevitable part of current and future business activity. 

Assignments in the AI era

 In light of this article from Radio NZ, whereby some universities in Aotearoa are no longer checking assessments using AI tracking platforms, a summary of ways to think about assessments in the AI age is of importance. There has been much discussion on how assessments in higher education need to be evaluated and re-thought, given the infiltration of AI into our work and study. This article in Times Higher Education, distills many of the main discussion points in academia on how AI affects academic writing.

The work undertaken at my institute is focused around holistic / programme wide assessment design, rather than on individual courses. The term 'programmatic assessments' is sometimes used to describe this approach

Some of the other strategies we have used, are summarised in this blog - NavigateAI (Dr. Ryan Baltrip)  In summary, to place greater weighting on recording the evidence of learning, rather than the product of learning. Therefore, portfolios and similar assessments are more useful than one off invigilated exams, or assignments. 

In Aotearoa, Otago Polytechnic's Bruno Balducci, have introduced the concept of AI safe design, a framework for the design of assessments which take into account the influences of AI. These are useful as a way to help educators work through the many pitfalls involved in redesigning assessments that will be authentic and relevant, but will not tempt learners into using AI to complete them.

The other concept we have used to help our teachers work out how to structure assessments in an AI age is the 'two lanes' assessment structure.  Here, lane 1 assessments are used to as assessments OF learning - or summative, higher stakes assessments. Lane two are the assessments FOR learning, taking on formative approaches to inform learners as they progress to the course.

Therefore, it is important to not just assume that current assessments will be appropriate but to undertake a stock take to understand the purposes of each assessment, and to put in place relevant assessments that will meet the purposes of each assessment i.e. evidence that the learner has met learning outcomes.  

Guide to using AI - school context

 Here is a useful guide for AI (in schools / US of A context). 

The guide begins with a section on how to use and why to use the guide.

The second section focuses on ethical issues - ensuring this is at the very front of any consideration for the use /integration of AI into teaching and learning. 

Discussions on the impact on students', risk and benefits and teacher perspectives follow.

The guide towards determining AI policies is then introduced and discussed. The 'how to create an AI policy' section is useful, drawing on key principles and providing suggestions. The checklist for developing AI policies (page 18) sets out the many parameters that need to be thought through as AI is introduced into the school curriculum.

A series of case studies and discussion pieces follow, documenting the struggles, challenges and pragmatic approaches adopted along with detailing various strategies and approaches. Discussions revolve around why, how, when and implications for introducing and using AI in schools. Strategies for assessments in the AI age are summarised (pages 30 -31) including the need to design engaging assessments, using paper based materials, having in-class assignments and assessments, adding oral assessments, emphasising the learning process, helping students understand the implications of using AI, clearly spelling out what is and what is not acceptable when using AI, and more frequent low stakes assignments.

A range of curated resources are provided for follow up and reference.

All in, a realistic documentation of how AI impacts on day to day school systems and environments. The pros and cons are drawn from case studies. The teacher voice comes through well and their perspectives and experiences are valued. Principles derived are relevant across educational sectors.