The Impact of Artificial Intelligence

Hangarau Mōhio — Work, Creativity, and Society

AI tools that can write, draw, code, and reason are no longer science fiction — they are here. Thinking clearly about their implications is now a core literacy for every student.

Key Concepts

Large Language Model (LLM)
An AI trained on vast amounts of text to predict and generate plausible language. ChatGPT, Claude, and Gemini are examples. They generate text, not understanding — an important distinction.

Automation bias
The tendency to over-trust automated or AI-generated information. Research shows people are less likely to question AI output than human output, even when the AI is wrong.

Algorithmic bias
AI systems trained on biased data reproduce and often amplify those biases. Systems trained primarily on English-language, Western data may perform poorly for Māori and Pacific users.

Data sovereignty
The right of communities to govern data about themselves. Māori data sovereignty (Tino Rangatiratanga o ngā raraunga) asserts that data about Māori should be controlled by Māori, not extracted by companies.

Comprehension Text

AI in the Classroom and the Workplace

In 2023, a New Zealand survey of secondary school teachers found that 78% had encountered student work that they suspected was AI-generated, but only 34% had a clear school policy on AI use. The rapid arrival of capable AI writing tools created a policy vacuum that schools are still working to fill.

The debate about AI in education mirrors a broader debate about AI and work. Economists disagree about whether AI will primarily destroy jobs, primarily create new ones, or primarily transform existing roles. Historical analogies are mixed: the industrial revolution eliminated many craft jobs while creating factory employment; personal computers eliminated typing pools while creating enormous new categories of work. Whether AI follows a similar pattern depends on political choices about redistribution, training, and investment — not just on the technology itself.

For Māori and Pacific communities, the implications are particularly complex. Many jobs historically accessible to people without tertiary qualifications — administration, data entry, customer service — are among those most threatened by AI automation. Meanwhile, the AI tools themselves are trained largely on English-language data, meaning their performance in te reo Māori is limited, and their understanding of tikanga and mātauranga Māori is shallow at best.

The Te Hiku Media group in Northland responded by building their own AI tools using recordings from the community — retaining data sovereignty and ensuring the technology served Māori language revitalisation rather than extracting from it. Their approach has been cited internationally as a model for indigenous communities navigating AI development.

Comprehension and Analysis Questions

1. Locate: According to the survey cited in the text, what percentage of teachers suspected AI use in student work? What percentage had a clear school policy?

2. Analyse: The text presents three different views on whether AI will affect employment. Summarise each view in one sentence.

3. Evaluate the example: How did Te Hiku Media's approach to AI differ from how most AI companies develop their tools? What principles does their approach embody?

4. Ethical reasoning: If an AI tool makes a decision that harms someone (e.g., incorrectly denies a benefit, misidentifies a person in a photo), who should be responsible — the company that built it, the organisation that used it, or no one? Explain your reasoning.

5. Argue (PEEL): Should AI-generated content be allowed in school assessments? Write a PEEL paragraph defending your position, considering both the learning implications and issues of fairness.

Whakaaro Hōhonu — Reflection

What is one skill you have that you believe AI cannot replicate? Why does it matter that humans retain that skill?

NZC Curriculum Alignment

Digital Technologies: Understand the social and ethical implications of digital technologies, including issues of data, privacy, and algorithmic decision-making.
English — Reading / Viewing: Think critically about how texts position readers, including the values embedded in technology narratives.

🌿 Te Ao Māori Lens

Te Hiku Media's approach reflects the principle of data sovereignty — that communities have the right to control knowledge about themselves. In te ao Māori, knowledge (mātauranga) is not extractable from the relationships and contexts in which it is embedded. AI systems that scrape data without consent or community partnership violate this principle. As an ākonga in Aotearoa, it is worth asking: when you use AI tools, whose data trained them, and who benefits from the value they create?