Introduction

"Our AI learning assistant summarises your course book, presentations, and live sessions and creates flashcards for you!"

Sound familiar? Many educational institutions are now promoting AI tools as magical solutions that can do the learning for you. This marketing pitch from one online university advertisement shows a concerning trend in education - the promise that AI can bypass the need for deep engagement with learning materials. Basically it’s telling students that learning is not relevant. Study progression is.

As we explored in our previous article, Learning with AI: Introduction, learning is a complex cognitive process that occurs in three crucial stages: encoding (initial processing of information), storage (retention in long-term memory), and retrieval (accessing stored information when needed). Each stage plays a vital role in building lasting knowledge and understanding. There are no shortcuts.

This is why promoting AI as a replacement for active learning is, frankly, educational malpractice. When we encourage students to let AI summarise their course materials and create their flashcards, we're essentially telling them to skip the encoding phase altogether. It's like trying to build muscle by watching someone else exercise - the results simply won't materialise.

Think about it: When an AI summarises a textbook chapter for you, you miss out on the crucial mental work of identifying key concepts, making connections to your existing knowledge, and wrestling with challenging ideas. When AI creates your flashcards, you lose the valuable cognitive processing that occurs while deciding what information is important enough to review later. These seemingly helpful shortcuts actually rob you of the deep processing that makes learning stick.

This doesn't mean AI has no place in education. On the contrary, when used thoughtfully, AI can be a powerful tool for enhancing our natural learning processes rather than replacing them. In this article, we'll focus on the encoding stage of learning - that critical first phase where new information is processed and prepared for long-term storage. We'll explore evidence-based strategies that students can use to improve their encoding process and examine how AI can support (not replace) these strategies.

Try This
Before we dive deeper into the encoding phase of learning, let's try a simple experiment inspired by Bower's research. This experience will show you firsthand how different encoding strategies affect your memory.

Now, cover the pairs above and try to complete each pair when given the first word:

• cheese - ?
• book - ?
• coffee - ?
• pencil - ?
• chair - ?

Now, let's try a different approach.

Cover these pairs and try to complete each pair when given the first word:

• cookie - ?
• tree - ?
• phone - ?
• lamp - ?
• piano - ?

How did your results compare between the two methods? Like Bower's participants, you likely remembered more pairs from the second set, where you created mental images. Perhaps you visualised a cookie being dropped into a mailbox, or a penguin using a lamp as a spotlight. These funny mental connections make the pairs more memorable than simple repetition.

Try to recall the pairs again tomorrow.

This simple exercise demonstrates a fundamental principle of effective encoding: when we actively engage with information by creating meaningful connections - even silly ones - we remember it better than when we merely repeat it. This is elaboration in action.

Understanding Encoding

Think of encoding as your brain's "save" function. When you encounter new information - whether reading a textbook, watching a lecture, or practicing a skill - your brain needs to process this information in a way that makes it stick. But, as we saw in the example above, not all encoding is created equal.

This is where elaboration comes in - a strategy that transforms superficial encounters with information into interconnected knowledge. Elaboration isn't just about repeating information; it's about making it meaningful by connecting it to what you already know, questioning it, and processing it deeply.

When you elaborate on information, you're essentially creating multiple pathways to access that knowledge in your brain. It's like building a network of roads to a destination instead of relying on a single path. The more connections you create, the easier it becomes to recall that information later.

Research has shown just how powerful these elaboration techniques can be. We briefly

described the study by Gordon Bower (1970) and creating mental images (1). In this study, participants who applied creating mental images remembered almost twice as much compared to those who simply repeated the words. When you visualise "piano-cigar," perhaps imagining a cartoon piano smoking a comically large cigar, you're not just memorising - you're creating a meaningful association that sticks.

Another helpful technique is called elaborative interrogation - asking and answering "how" and "why" questions about what you're learning.

What happens when students ask "why" questions while studying? A study with middle school students showed us just how powerful this simple technique can be.

The researchers (2) worked with students in grades 6 and 7, giving them various facts to learn. Some facts aligned with what students already knew, while others were surprising or contradicted their expectations. The students were divided into different groups to test three study methods:

1. Asking and answering "why" questions about the facts (elaborative interrogation)
2. Reading the facts and trying to understand them
3. Using whatever study method they preferred

Students were tested on these facts immediately after studying, a month later, and two months later. The results were clear: students who used elaborative interrogation - asking "why" questions - remembered significantly more than those who simply read for understanding or used their own study methods.

Interestingly, it didn't matter whether students studied alone or with a partner. What did matter was the quality of their "why" explanations. When students came up with good explanations that helped clarify the facts (or heard such explanations from their study partners), they remembered those facts much better.

This research shows us something powerful: simply asking "Why is this true?" and attempting to explain it to yourself can dramatically improve your learning, regardless of whether you study alone or with others.

The key insight here is that effective encoding isn't passive. You can't just expose yourself to information and expect it to stick - that is, you can't ask AI to do the work for you. Having ChatGPT, Claude, or some other tool summarise a textbook chapter might save time, but it bypasses the crucial encoding process that makes learning stick. The fact that over 58% of students don’t feel equipped to use AI-tools (3) emphasises the need to ensure students understand how AI can support and improve their learning and not replace it.

AI Tools Supporting the Encoding Phase

ChatGPT has emerged as students' preferred AI tool for learning support, followed by other popular applications like Grammarly, Microsoft Copilot, Google Gemini, and Perplexity (3). However, the key to effectively using these tools lies in leveraging them to enhance, rather than replace, active learning techniques.

Techniques mostly used when learning from texts:
note-taking, summarising, grouping, forming questions (4),
elaborative interrogation, self-explanation, highlighting and underlining (5).

So, instead of asking AI to complete tasks, you can use these tools to deepen your understanding, challenge your thinking, and develop more sophisticated learning approaches. For example, rather than requesting direct summaries or notes, you can engage with AI to analyse your own summaries, enhance note-taking techniques, develop more effective grouping strategies, and create thought-provoking questions. This approach ensures that AI serves as a learning scaffold that promotes deeper cognitive processing during the encoding phase.

Here are a few ideas on how to prompt your AI tool instead of summarise this text or create flashcards for this chapter.

Learning technique: Note-taking
Description: Recording key information while reading text
Prompt examples:
1) "I've taken these notes from Chapter 3 on [topic]: [paste notes]. Can you review my note-taking approach and suggest ways to better organise the information to show relationships between concepts?"
2) "Based on these notes I took [paste notes], check if I’m missing any key information?"

Learning technique: Summarising
Description: Condensing text to capture main ideas
Prompt examples:
1) "Here's my summary of the [topic]: [paste summary]. Can you help me identify any gaps in my reasoning or important connections I might have missed?"
2) "Here are two summaries I wrote of the same text. Which one better captures the key concepts and why?"

Learning technique: Grouping/Categorising
Description: Organising information into meaningful categories
Prompt examples:
1) "I've grouped these historical events into these categories: [paste groupings]. Can you suggest alternative ways to organise this information?"
2) "I've created these categories: [paste]. What concepts might fit into multiple categories and why?"

Learning technique: Question formation
Description: Creating questions to deepen understanding
Prompt examples:
1) "For this text about climate change [paste excerpt], I've created these questions: [paste questions]. How could I modify them for deeper analysis?"
2) "I'm trying to create questions that connect different chapters. Are these questions effectively bridging the concepts?"

Learning technique: Elaborative interrogation
Description: Generating "why" questions and answers to promote understanding
Prompt examples:
1) "I've created these why-questions for the concept of natural selection: [paste questions and answers]. Can you suggest areas where I could deepen my reasoning?"  
2) "I've written these why-questions and explanations. Does any need more specific evidence or examples?"

Learning technique: Self-explanation
Description: Explaining concepts in your own words while learning
Prompt examples:
1) "Here's how I've explained the concept of supply and demand to myself: [paste explanation]. Are there any misconceptions or gaps in my understanding?"
2) "Here's my explanation of [concept]. What analogies might make my explanation clearer?"

Learning technique: Highlighting/Underlining
Description: Marking important information in text
Prompt examples:
1) "I've highlighted these parts of the text as key points: [paste highlighted sections]. Can you help me evaluate if these are essential or supporting details?"
2) "Here's what I highlighted as key points versus supporting details. Is my categorisation logical?"

The key is that each prompt asks the AI to evaluate, enhance, or provide feedback on the student's own work, rather than asking it to do the work itself. This approach maintains active engagement while using AI as a tool for improvement and deeper understanding.

Conclusion

AI tools are powerful in education. But that’s what they are and that’s how we should promote them – tools!

Rather than using AI to bypass active learning, students should leverage these technologies to enhance their existing learning strategies and deepen their understanding. The key is to have students complete their own work first, then using AI for feedback, refinement, and metacognitive development.

The prompts and strategies outlined above demonstrate how AI can serve as a cognitive partner rather than a replacement for student effort. When used thoughtfully, AI tools can help students identify gaps in their understanding, make new connections between concepts, and develop more sophisticated learning approaches. None of this can be achieved if the tool summarises texts and presentations, takes notes or creates flashcards instead of students.

As AI continues to evolve, it is our responsibility as educators to keep the focus on promoting deep learning rather than surface-level task completion. By using AI to enhance rather than replace traditional elaboration strategies, students can develop stronger learning skills while building a deeper understanding of their subject matter. The future of AI in education should be supporting and expanding human learning capabilities and NOT replacing them.

Read more in:

1) Bower, G.H., Winzenz, D. (1970). Comparison of associative learning strategies. Psychon Sci 20, 119–120. https://doi.org/10.3758/BF03335632 

2) Woloshyn, V. E., & Stockley, D. B. (1995). Helping students acquire belief-inconsistent and belief-consistent science facts: Comparisons between individual and dyad study using elaborative interrogation self-selected study and repetitious-reading. Applied Cognitive Psychology, 9, 75-89.

3) https://universitybusiness.com/these-5-ai-tools-are-the-most-popular-among-students/

4) Tay, B. (2013). Elaboration and organization strategies used by prospective class teachers while studying social studies education textbooks. Egitim Arastirmalari-Euroasia Journal of Educational Research, 51, 229-252. (Link: https://files.eric.ed.gov/fulltext/EJ1059832.pdf)

5) Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4-58. https://doi.org/10.1177/1529100612453266