Beyond Infographics: How to Use Nano Banana to *Actually* Support Learning
Six evidence-based use cases to try in Google's latest image-generating AI tool
Hello folks 👋
Over the last couple of weeks, there’s been a lot of buzz about Nano Banana — Google’s latest AI image generation tool.
Most of the conversation among educators has focused on Nano Banana’s ability to create infographics with polished, labelled, on-brand visuals. While it’s true that Nano Banana generates better infographics than other AI models, the conversation about this technology has so far massively under-sold what’s actually different and valuable about this tool for those of us who design learning experiences.
While it’s true that Nano Banana generates better infographics than other AI models, the conversation has so far massively under-sold what’s actually different and valuable about this tool for those of us who design learning experiences.
So in this week’s post, I’m breaking down six of the most research-supported learning strategies from cognitive and educational science — and showing exactly how Nano Banana can operationalise each one.
Each section includes a research-based summary of what the strategy is and why it’s powerful, how Nano Banana uniquely supports it, and a dynamic prompt you can paste into your own workflow.
Let’s dive in!
Nano Banana 101
Before we dive into the playbook, here’s a quick overview of makes Nano Banana different from other AI image generation tools:
Most commercial AI image generators (Midjourney, DALL-E, Stable Diffusion) are trained on massive, unfiltered internet image collections. They can excel at photorealism and aesthetic novelty, but they have significant limitations for instructional design:
Training data bias: They over-index on stock photography aesthetics, which often means generic, disconnected, or non-diverse representations. When you ask for “a student studying,” you get the same corporate-training-room visual millions of people have seen.
No sequential memory: Each prompt is independent. There’s no way to ensure the same character, colour scheme, or visual logic persists across 10 images. Consistency—which is essential for storyboards, worked examples, and contrasting cases—requires manual post-production or luck.
No instructional awareness: They don’t understand learning design terminology (generation effect, scaffolding, misconception, worked example). You have to translate your pedagogical intent into generic visual language, which often gets lost.
No multimodal input integration: They accept text only. You can’t feed them your rough sketch, your brand guide, your learner demographics, or your previous output and ask the model to build on it. Each image is a fresh start.
In controlled testing, Nano Banana’s unique capabilities suggest a superior model architecture optimised for deeply accurate visual understanding and multimodal reasoning across complex, real-world scenarios (Truong et al, 2025).
In practice, what makes Nano Banana unique is:
Multimodal inputs: Sketches, screenshots, photos, existing diagrams, and brand assets feed into the model, shaping every new output. Your rough idea becomes the scaffolding for the refined visual.
Persistent visual identity: You can lock characters, palettes, layouts, and visual language across sequences. The model understands “use this character in all 6 panels” and “keep the same office setting but vary the interaction.” better than any other.
Controlled variation: You can systematically change specific parameters within an image while holding everything else constant, making it easier to edit the images it produces and ensure their quality.
What this means for our workflow:
Instead of the traditional “commission → wait → tweak → approve → repeat” cycle, Nano Banana enables an iterative, rapid-cycle design process where you can:
Sketch an idea and see it refined in minutes.
Test multiple visual metaphors for the same concept without re-briefing a designer.
Build 10-image storyboards with perfect consistency by specifying the constraints once, not manually editing each frame.
Implement evidence-based strategies (contrasting cases, worked examples, observational learning) that are usually too labour-intensive to produce at scale.
This shift—from “image generation as decoration” to “image generation as instructional scaffolding”—is what makes Nano Banana uniquely useful for the 10 evidence-based strategies below.
Six Nana Banana Use Cases to Drive Better Learning Outcomes
1. Visualisation
What it is
Visualisation means creating a spatial representation—like a diagram, map, timeline, or path diagram—to organise ideas and information outside the head. Work on external representations shows that when information is laid out in space, people can search it, see patterns, and notice structure more easily, which in turn supports better problem solving and learning (Schwartz, 1993; Zhang, 1997).
Studies of drawing-to-learn in science also find that asking students to sketch systems and data improves engagement, representation, and scientific reasoning, and that spatial skills themselves can be improved with practice (Ainsworth, Prain, & Tytler, 2011; Uttal et al., 2013).
Why it’s powerful
Visualisation offloads working memory by making structure visible, and it highlights patterns and causal connections.
How Nano Banana helps
Converts text-heavy explanations into crisp, structured diagrams
Harmonises multiple messy inputs (slides, notes, docs) into one coherent visualisation
Represents invisible processes (algorithms, systems, pathways) that need spatial structure
Example Prompt to Try
“Visualisation organises information spatially, reducing working-memory load and making structural relationships easier to grasp. Research shows that well-structured diagrams, maps, and flows improve comprehension because they externalise cognitive structure, allowing learners to “see” relationships that would otherwise need to be mentally juggled. Using these process notes and workflow screenshots, generate a structured visualisation (flowchart, swim lane map, or system diagram) that clarifies relationships, dependencies, and sequencing.”
2. Analogy
What it is
Analogy-based learning involves identifying shared deep structure between two superficially different examples. Decades of research show that analogies help learners notice underlying principles and transfer them to new contexts — from preschool transfer learning (Brown & Kane, 1988) to classic studies of schema induction (Gick & Holyoak, 1983) and expert conceptual change (Gentner et al., 1997).
Students learn most when they compare multiple examples, because variability reveals what truly matters (Richland, Zur, & Holyoak, 2007).
Why it’s powerful
Analogies help learners see deep structure rather than surface details, and they greatly improve transfer to unfamiliar problems.
How Nano Banana helps
Instantly generates parallel visual metaphors with controlled structural alignment
Lets you test and refine alternative analogies quickly
Produces diagrams that hold deep structure constant while varying context
Example Prompt to Try
“Analogy-based learning involves identifying shared deep structure between two superficially different examples. Using the relevant parts of this course storyboard, generate a side-by-side diagram explaining X using the analogy of Y. Ensure each element of X maps clearly to its counterpart in Y. Maintain consistent colour coding and aligned spatial structure to emphasise the shared deep principle. Never use the work “analogy” in the image.”
3. Worked Examples
What it is
Worked examples are fully solved problems that make an expert’s steps and reasoning visible. Research comparing worked-example study with unguided problem solving shows that beginners often learn procedures more accurately and efficiently when they follow a clear, step-by-step solution instead of having to invent every move themselves, because this reduces unnecessary cognitive load (Sweller, 1994; Atkinson, Derry, Renkl, & Wortham, 2000).
Follow-up work has shown that well-designed examples help learners grasp useful subgoals and patterns, especially when combined with prompts to explain each step (Catrambone & Holyoak, 1990; Renkl, Stark, Gruber, & Mandl, 1998).
Why it’s powerful
They reduce cognitive load for beginners and make invisible reasoning steps explicit.
How Nano Banana helps
Converts multi-step processes into clean visual sequences
Enables rapid production of multiple worked examples with structural variations
Supports faded examples by gradually removing visual scaffolds
Example Prompt to Try
“Worked examples are fully solved problems that make expert thinking visible. Research on cognitive load shows that novices learn more efficiently when they study explicit steps rather than solving everything from scratch. Using the relevant parts of this course storyboard, create a 5-step visual worked example illustrating the solution process. Ensure each step highlights key information, with consistent visual scaffolding that fades in later steps.”
4. Contrasting Cases
What it is
Contrasting cases are sets of examples that appear on first glance to be the same but differ on the key feature which learners must notice. Research shows that juxtaposition and close comparison sharpens perceptual discrimination by helping learners extract structure — a foundational aspect of expertise seen in domains from wine tasting (Solomon, 1990) to archeology (Goodwin, 1994).
Carefully varied examples accelerate structure detection and subsequent transfer, as demonstrated in mathematics perceptual learning modules (Kellman, Massey, & Son, 2010).
Why it’s powerful
Contrasting cases help novices detect subtle yet critical differences, enabling better categorisation and more accurate decision-making.
How Nano Banana helps
Generates controlled families of images with fixed deep structure
Allows systematic variation of context, risk factors, or features
Makes iteration instant, allowing fast experimentation
Example Prompt to Try
“Contrasting cases are sets of examples that differ only on the key feature learners must notice. Research shows that juxtaposition and close comparison sharpens perceptual discrimination by helping learners extract structure. Using this scenario brief, create a 2×2 grid of contrasting cases that hold layout and angle constant while varying only the predefined critical features. Label each case using the categories from the brief to highlight the intended distinctions.”
5. Elaboration
What it is
Elaboration requires learners to expand on a topic — for example, by explaining it in your own words, linking it to a story, or imagining a concrete example. Classic studies show that when learners create relevant and specific elaborations, they remember much more later than when they only repeat or copy information (Stein & Bransford, 1979; Tresselt & Mayzner, 1960).
Even young children can be taught simple elaboration strategies, like imagining two things interacting in a vivid way to remember them as a pair (Yuille & Catchpole, 1973).
Why it’s powerful
Elaboration deepens understanding by creating densely connected knowledge structures, and it improves cued recall and flexible use of knowledge.
How Nano Banana helps
Generates context-rich scenes grounded in learners’ environments
Produces multiple examples around a single idea (different roles, settings, stakeholders)
Converts learners’ own stories or sketches into polished illustrations for elaborative anchoring
Example Prompt to Try
“Elaboration strengthens memory by connecting new information to prior knowledge, making it richer and more retrievable. Using the use cases in this training outline, create three varied scenes that elaborate the core concept by embedding it in different everyday contexts. Ensure each scene reflects a distinct stakeholder or environment from the outline.”
6. Generation
What it is
Generation requires learners to produce an answer, explanation, or prediction before seeing the correct one. Research shows that generating — even incorrectly — produces stronger memory than passive exposure (Slamecka & Graf, 1978).
Retrieval-focused generation also enhances long-term retention (Bjork, 1994), outperforming elaboration in some learning tasks (Karpicke & Blunt, 2011). Generation works by strengthening retrieval pathways through active reconstruction.
Why it’s powerful
It improves retention and understanding through effortful retrieval, and it reveals misconceptions early.
How Nano Banana helps
Creates intentionally incomplete diagrams, scenes, or flows
Supports stepwise “guess-then-check” sequences
Generates multiple isomorphic prompts to encourage repeated retrieval
Example Prompt to Try
“Generation requires learners to produce an answer, explanation, or prediction before seeing the correct one. Research shows that generating produces stronger memory than passive exposure. Using this module outline, create an unlabelled or partially revealed diagram of X that requires learners to generate missing steps or labels. Leave intentional gaps aligned to the module’s learning objectives.”
Closing Thoughts
The key message here is that, when used with intention by a specialist, AI tools have the potential to become powerful engines to enhance the value and impact of the learning experience.
These six strategies — grounded in decades of robust research — represent some of the most powerful tools we have for improving learning. Nano Banana’s unique multimodal capabilities make these strategies not only accessible but scalable in your everyday workflow.
The key message here is: AI needs you! Only when you pair AI tools with proven learning mechanics, do our results elevate from prettier content to better learning. To truly realise the power of AI, we must concentrate first on a deep understanding of how humans learn and only then ask — can this AI tool help to operationalise it?
Only when you pair AI tools with proven learning mechanics, do our results elevate from prettier content to better learning.
Happy innovating! Phil 👋
PS: Want to dive deeper into how to use AI to 10X the value and impact of human learning? Check out my AI & Learning Design Bootcamp where we explore exactly how to operationalise learning science with emerging technologies.
