How AI is transforming student learning

Whether preparing for exams or diving deep into complex concepts, artificial intelligence (AI) is increasingly being used in higher education to support learning. Students are using it in many potentially helpful ways, from consolidating lecture notes to acting as a personal tutor. GenAI has evolved far beyond simple question-and-answer tools to act as an interactive learning partner that's available around the clock. 

Dr Joshua Pate, senior lecturer in physiotherapy and researcher at UTS, is leading AI research in education with colleagues at the Faculty of Health. He's seen firsthand how AI is changing the learning landscape, and he has some important insights about making the most of these powerful tools while avoiding potential pitfalls. 

Here's what students needs to know about learning with AI.

From search engine to study buddy

Gone are the days when AI was considered to be some sort of fancy search engine. Research shows that today's students are using AI for wide range of academic tasks, including summarising notes, clarifying concepts and creating study materials personalised to their needs. In this sense, it can be like having a study buddy who never gets tired, is available 24/7 and doesn't cost anything extra. 

“Even before the latest GenAI developments, students were already moving beyond quick answers by using AI to challenge their thinking and consolidate their own notes,” explains Pate. 

This shift from passive information gathering to active engagement is helping students interact with their educational content in completely new ways. 

For example, GenAI chatbots can adjust explanations based on how well you understand something. Struggling with a complex concept? Ask it to break things down more simply. Need a challenge? Request harder questions or counter-arguments to test your thinking. This scaffolded approach means you can learn at your own pace and in your own style. 

“AI can enhance learning when students already have a foundation in a subject. With context, it can test knowledge, challenge assumptions and help students apply skills across different contexts.” 

GenAI systems can also be effective for brainstorming and exploring different perspectives. Imagine you're working on a project about healthcare policy. AI can potentially simulate discussions between different stakeholders, from school principals to health professionals and even the World Health Organisation, giving you rich insights into how different groups might approach the same complex issue. 

Why working smarter doesn't mean working less

Here's where things get tricky. While AI offers incredible possible benefits, there are some serious risks to consider. 

As Pate points out: “Learning new things requires effort, and so if people aren't putting in any effort, then they may not be learning.” 

Over-relying on AI may also instil a false sense of confidence. This is increasingly important as ‘hallucination’ rates in AI chatbot models, although reduced, are still occurring. You might think you understand something because AI seemingly explained it clearly, but can you actually apply that knowledge when it matters? There's also the risk of missing critical thinking skills or becoming too dependent on AI to spot errors in information. These risks alone have important implications for assessment tasks. 

GenAI also has a tendency to be very agreeable, leaning toward validating your ideas and telling you what you want to hear. 

“There's a risk of students being told (inaccurately) they're right all the time. This risk has drastically reduced, but it's still a big risk," warns Pate. 

“This could undermine resilience, critical thinking and the ability to learn from mistakes.” 

The most advanced AI systems aren't perfect. On one hand, chatbots can now get a gold medal in a maths olympiad, but at the same time, they can struggle with some primary school level questions. They can make mistakes, provide biased information, and as Pate puts it, they follow the "garbage in, garbage out" principle. If the data AI was trained on wasn't accurate, its outputs won't be either. 

Why you still need to master the basics 

Here's something crucial: you need solid foundational knowledge to make the most of AI. Without understanding the core content of your subject, you can't properly evaluate whether AI's responses are accurate, complete or relevant. 

This is especially important in fields like healthcare, where misunderstanding AI advice could have serious consequences. But it applies to all disciplines, because you need that foundational knowledge to think critically about what AI tells you. 

Building this foundation could enhance what experts call "metacognitive skills" – essentially, learning how to learn. These skills are important for your brain health and support lifelong learning and help you adapt as your field evolves.