AI, particularly generative AI, has flooded the learning space. While some AI tools are undeniably useful, others may be questionable. Even as the NCEA itself is rolling out AI tools to assist with exam marking, and adapt learning materials to suit learner needs and preferences, questions remain about the ability of AI to be used as a plagiarism tool, reducing the ultimate goal of education – students improving their knowledge and skills. But it is impossible to remove all access to AI, even if we wanted to.
So what can we do, as parents and educators, to help our children navigate this tricky arena?
LLMs and Fact-checking
Generative AI is present in so many digital apps, including Microsoft Copilot, Grammarly AI, Google AI Overview and more, that they have become a popular option to find information. Despite the huge amount of data that these tools are trained on, they can still hallucinate or provide strings of text that don’t accurately represent the information asked for. Students can think they are finding the right answers when in fact they may not be.
In order to overcome this, we still need our ability to evaluate and navigate the answers AI provides us. Increasingly, AI tools such as Perplexity and Google AI Overview provide links to the websites where they found their information. This allows us to cross-check the information and make sure that the sources are authoritative.
Cheating and Plagiarism
A major concern for assessments is academic honesty. When assessments require students to create texts such as reports, essays, coding and more as proof of their skills, it is possible for many students to prompt an AI tool to create this text for them. While most experienced students understand that this is dishonest, they may be tempted to take shortcuts if they are under pressure. Even honest students can also use AI tools to assist their work, in the belief that this is acceptable because they participated in the work directly too.
Given how prevalent AI tools are, including in services like spelling and grammar checking, identifying the reason why using them is acceptable or unacceptable is vital.
The AI Detector and the AI Humanizer
Many teachers and markers use a range of AI detectors to identify where students have used AI to an unacceptable level. Unfortunately, AI detectors cannot be 100% accurate in their results, and AI tools designed to combat AI detectors also exist. Therefore, simply identifying a student’s use of AI is not sufficient to address these issues.
Many schools and universities are developing an AI use policy. At this stage, some schools state a specific percentage of AI results that is permissible. As AI detectors do not return the same results on the same texts, and are not reliable, it is not feasible to require a score of 0% from AI detectors. For example, if a school states that all submitted essays must score as AI written less than 25%, this gives a clear standard students can check if they have reached.
Cognitive Offloading
In addition to academic dishonesty, a student’s education can also be compromised if they are ‘outsourcing’ their thinking to an AI tool. If a student does not have sufficient background knowledge or skills, they will not improve these by disengaging with the learning process.
Teachers and education experts know that whatever tools and technology is used, a student must be engaged in cognitive effort in order to learn. This does not mean that all effort is equal. (See our previous posts about cognitive load theory here.) While some kinds of knowledge (such as some rote memorization) may not be essential in today’s world, many kinds of knowledge and skills are still vital, particularly in order to assess and improve the output of technology.
Next Steps
AI technology is here to stay, and it will most likely form part of our working lives in the future. As parents and educators, we need to make sure our students are improving their AI literacy skills, so that they can use the technology appropriately, in an honest, productive way.
Students who are engaged, understand the limits and value of AI, and think critically and creatively about their use of the technology, are well-placed to enter further study, the workforce, or pursue their aspirations.
Resources
New Study Says AI is Making Us Stupid But Does it Have To? (Forbes)
Increased AI Use Linked to Eroding Critical Thinking Skills (Phys.org)
AI and Cognitive Offloading – Sharing the Thinking Process with Machines (UX Collective / Medium)
Artificial Intelligence Exam-marking on the Way for Year 10 Writing Tests (RNZ)
Leave a Reply