Using AI-Enabled Images and Comic Narrative to Enhance Entrepreneurship Education

Abstract

This learning innovation is an engaging technique to help students learn abstract concepts such as entrepreneurship theory, mindsets, and methods in a playful and exciting way. We explore how current AI tools can be used by entrepreneurship educators to enhance their course materials by creating memorable images and story-based, highly visual comic book content (“graphic novels”). We implemented these enhancements in an interactive learning platform based on Articulate Storyline and tested them in an introductory entrepreneurship course with over 2000 students per year on building an entrepreneurial mindset. In this article, we include sample copies of the images and graphic novel comics and highlight key methods, lessons, results, and design artifacts created during the course of our learning journey. Educators may add these free Open Educational Resource (OER) assets to their own courses by linking to the online version. We believe this approach to entrepreneurship education enhances students’ engagement, memory retention, and the ability to identify and name underlying concepts and entrepreneurship theories. We hope to spark the interest of other entrepreneurship educators to use similar approaches at their universities.

Theoretical Background

It is widely known that university students don’t read as much as they used to, and getting them to engage with course textbooks and instructional materials is increasingly difficult (St Clair-Thompson et al., 2018). As many as three out of four university students report that they rarely or never complete their reading assignments on schedule (Connor-Greene, 2000). Many potential reasons have been given for this lack of reading, including lack of motivation (Hatteberg & Steffy, 2013), poor attitudes towards reading (Marek & Christopher, 2011), low levels of reading comprehension (Ryan, 2006), lack of confidence with reading (Lei et al., 2010), and students’ lack of time (Hoeft, 2012).

Asking students to watch videos as an alternative to reading is also challenging. In one of the author’s university courses, a required introductory course for over 2000 first-year business students per year, fewer than half the students will click on the required course videos at all, and fewer than half of these will watch beyond the first 30 seconds. Only 20% of students will watch any given 5–10-min instructional video until the end. In other words, only 400 out of 2000 students will watch any given high-quality video assigned—about the same percent as will read any given assignment (e.g., Connor-Greene, 2000). Analytics performed on the author’s YouTube channel indicated that these viewer statistics are “average” for all YouTube videos.

Comic books and graphic novels may provide educational merit as a useful supplement to traditional text-only reading by appealing to students through their multimodal engagement (Krusemark, 2017). Comics can be defined as a storytelling medium that uses text and sequential art, that is, images used in a specific order (Eisner, 2008). Comic images form an intimate part of the storytelling and allow the reader to decode, understand, navigate, and interpret written language, visual images, and the design and relationship between these qualities (Serafini, 2011). Short et al. (2013) found that long-form narrative comics (i.e., “graphic novels”) were particularly useful for improving comprehension, retention, and engagement over traditional format textbooks.

We draw inspiration, in particular, from the 2008 book “The Adventures of Johnny Bunko” by Daniel Pink (2008). Pink writes engaging business books that are frequently featured on best-seller lists. However, he also wanted to connect with younger readers who might benefit from early career advice. Instead of writing yet another traditional self-help book, he teamed up with the illustrator Rob Ten Pas. Together, they packaged six essential principles for a successful career and created a highly engaging comic book with a coherent story that allows readers to learn about these principles in a more narrative form.

Unfortunately, few entrepreneurship educators have the time and resources necessary to ideate and write a long-form comic narrative story, hire an illustrator, and print and package a graphic novel of this quality for their courses. This process of turning abstract concepts such as entrepreneurship theory, mindsets, and methods into narrative storylines and subsequently creating complementary visuals is one of the many use cases where current AI tools might allow us unprecedented creative opportunities.

We wanted to see if current AI applications such as ChatGPT, Bard, DALL-E, and Midjourney could facilitate a significantly broadened spectrum of creative expression that we—as playful entrepreneurship educators—could use to enhance our students’ engagement, memory retention, and ability to learn entrepreneurial concepts.

We want to take up Winkler et al.'s (2023, p. 579) invitation to “embrace this transformative [AI] innovation to not only educate our students but also use its potential to reshape our classrooms,” which was recently published as an editorial in Entrepreneurship Education and Pedagogy. Although taking the first steps with different AI tools might seem daunting at first, the impact these technologies will have on teaching practices clearly highlights the need to explore this space.

Researchers have been discussing the impact of AI tools on entrepreneurship education in general (Bell & Bell, 2023), the paradoxes of its use in management education (Lim et al., 2023), the varied effects on creative destruction (Norbäck & Persson, 2023), how AI might impact dealing with uncertainty in entrepreneurial decision making (Townsend & Hunt, 2019), the impact on entrepreneurial career intentions (Park & Sung, 2023), its ability to personalize learning for students (Baidu-Anu & Ansah, 2023; Su & Yang, 2023), how to utilize AI tools to create more value for entrepreneurship students (Darnell & Gopalkrishnan, 2024), and how we as a community of educators might cope with these recent developments (Mills et al., 2023).

Among the current stream of publications in this field, we also found exciting and uncommon research approaches such as the one employed by Michel-Villarreal et al. (2023, p. 1), who have used an ethnography approach to frame their paper on the “challenges and opportunities of generative AI for higher education as explained by ChatGPT.”

Design Science Research (DSR) provides an important approach to advancing and sharing knowledge about the use of AI to enhance education through the creation of innovative artifacts (vom Brocke et al., 2020). DSR uses the scientific method to improve design practice, create artifacts that capture long-lasting knowledge, and connect research and practice (Telenko et al., 2015). Unlike descriptive science, which deals with historical facts to be interpreted, design science is primarily concerned with how to design future possibilities (Dimov, 2020) and is thus a vital approach to improving the design, delivery, and performance of learning innovations.

We apply a DSR approach to our learning innovation and provide a sample of how we integrated AI tools from ChatGPT and Midjourney into our interactive learning platform authored using RISE 360 and published as an Open Educational Resource (OER) using Articulate Storyline. We provide and describe how to use the learning innovation in PowerPoint slides and course materials and deploy it through the university’s Learning Management System (LMS).

What is the Innovation?

We created AI-enabled images and comic book narrative stories (a “graphic novel”) embedded into an interactive web 3.0 learning platform accessed through our university’s LMS. Sample multimodal images are shown in Figures 1–7 (which can also be used as PowerPoint slides in class). Design artifacts for the images and text of the first three chapters of a graphic novel are available in the Supplemental Material section on the journal website, along with sample knowledge checks to ensure comprehension. The interactive web 3.0 OER version may be found at https://entrepreneurinstitute.ca/building-an-entrepreneurial-mindset-personal-focus (Gedeon, 2024).

Figure 1.

AI image representing today’s mental health crisis.

Figure 2.

AI image representing the need to learn to use AI.

Figure 3.

AI image representing growth mindset.

Figure 4.

AI image representing the entrepreneurial mindset of curiosity.

Figure 5.

Alternative AI image representing the entrepreneurial mindset of curiosity

Figure 6.

AI image representing the entrepreneurial mindset of self-leadership.

Figure 7.

Alternative AI image representing the entrepreneurial mindset of self-leadership.

This is a demonstration of the current state-of-the-art (Fall 2024) of what AI tools can (and cannot yet) produce as graphic novels when used by non-experts (i.e., university entrepreneurship educators with limited time and resources). Our goal is to encourage other professors and show what is currently feasible for them to create on their own. This approach role-models the behaviour we want to see in our students. We aim to demonstrate that we, as educators, are engaged, curious, playful, and creative. We hope this inspires others to try something new.

For Whom is the Learning Innovation Intended?

The learning innovation was primarily created for an introductory entrepreneurship course about building an entrepreneurial mindset with over 2000 first-year business students as well as non-business students at a large Canadian university. For most students, this is their first university course. Most students will be under 20 years old.

Students spend 1 to 2 hours per week engaging with the course’s interactive Articulate Storyline learning system accessed through the university’s LMS (D2L Brightspace). Students progress through the learning system by reading, watching videos, successfully completing all interactive learning objects and comprehension tasks, and passing quizzes. These learning activities also help students successfully pass each week’s graded assignments. The interactive graphic novel chapters appear as one lesson out of ten in the course’s weekly modules.

What are the Expected Learning Outcomes?

We expect that the AI-generated images (such as shown in Figures 1–7) will enhance student retention of key entrepreneurial concepts. This is particularly useful since the course contains a lot of material, and we want to highlight certain core concepts for long-term retention so they don’t get lost amidst all the other course content. Mental images for many learning types are key to memory retention.

We expect that comic narrative storytelling in graphic novels will concretize abstract entrepreneurial concepts and help students identify and reinforce key entrepreneurial ideas and tools including:

• Creativity and great ideas don’t normally happen on their own.

• User-centric and iterative design thinking methods work better than trying to start with a great idea and then designing a product or service all on your own.

• Effectuation principles like the Bird-in-the-Hand and Affordable Loss.

• You will encounter many challenges on your entrepreneurial journey. These are not just normal; they are a GOOD THING and provide barriers to entry as well as growth opportunities.

• Proper and improper examples of how to use tools like the Purpose Statement and Customer Persona.

• Deciding when and where to pivot.

• Monitoring your thoughts to identify Negative Self-Talk and Cognitive Distortions.

• Re-Framing as a way to build Resiliency and a Growth Mindset.

We expect that using multimodal and interactive learning environments with heroic themes and terminology related to students growing their superpowers will enhance our students’ engagement and interest in entrepreneurship. We expect them to be able to see themselves in the characters we are portraying and aspire to grow and learn along with these characters.

How Does it Work?

We started our journey by playing with the major end-user AI platforms, reading articles, and watching tutorials on YouTube. We tested all the comic book generation websites, such as ComicsMaker, Comic Factory, Comicai, and Stable Diffusion, but ultimately found that ChatGPT (version GPT-4 turbo) was the most useful tool for narrative storylines and Midjourney (version 6.1) was the most effective image-generation tool. We cut and pasted the best images into PowerPoint where we added text, speaking bubbles, and comic book panels as desired.

We were able to quickly use AI-generated images to improve our in-class lecture slides, and we provide representative design artifacts in Figures 1–7. Figure 1 uses rich graphics to represent emotional prompt terms such as sadness, grief, anxiety, and depression. Figure 2 is also highly memorable and conveys the fear of being replaced by AI if students do not learn to use it.

Each of the course’s core entrepreneurial attitudes was given its own memorable representative image to reinforce that abstract idea throughout the course. This was an important learning innovation because the course introduces a wide variety of topics, including change and uncertainty, the future of work, searching versus planning, effectuation, design thinking, lean startup, business model canvas, strategy, the business plan, company lifecycle, goal-setting, habit formation, happiness, well-being, Cognitive Behavioral Therapy, positive psychology, and human, social, and financial capital. With all that subject matter to cover in a 13-week introductory course, it is easy for students to lose track of the key theme of the course, which is building an entrepreneurial mindset by building curiosity, proactivity, resiliency, growth mindset, creativity, empathy, optimism, self-leadership, self-esteem, and self-efficacy (each of which now has its own distinctive imagery associated with it throughout the course content and in slides).

Figure 3 shows a sample image used for the attitude of growth mindset. Figures 4 and 5 depict sample images used for curiosity. Figures 6 and 7 are used to represent the attitude of self-leadership, which is based on using your conscious System 2 Executive to control your subconscious System 1 Elephant (Kahneman, 2011). We created dozens of slight variations of these images to reinforce these key entrepreneurial mindsets throughout our instructional materials.

Our next task was to string together a series of images along with a good narrative storyline to create comic book chapters that demonstrate entrepreneurial mindset concepts and the use of tools. We started by following the prompting strategies of Schulhoff et al. (2024) to create graphic novel comic narrative scripts and scene descriptions. Our design criteria included the following :

• An engaging main character who is a university student with an unknown “superpower.”

• An engaging story about starting up a new venture.

• An interesting venture.

• The characters use various design thinking tools and methods such as interviewing, prototyping, purpose statement, customer persona, and customer profile canvas.

• The characters demonstrate entrepreneurial mindsets like proactivity, curiosity, resiliency, and growth mindset.

• The characters sometimes demonstrate negative self-talk and cognitive distortions as well as use of positive self-talk.

• The protagonist interacts with and starts forming a new venture team with other characters.

• The characters go through various adventures and venture pivots.

We were unable to generate any single script that met all these design criteria using AI. However, we were inspired by enough portions of the 20 to 40 ChatGPT response iterations to at least drive the image generation process.

We then used Midjourney to create scenes and characters. At present, it is still difficult to get AI image generators to create consistent characters across various prompts, even if you provide visual source material—either by uploading reference images or asking the AI tool to use specific reference material, such as a series of existing comic books by the same illustrators. Midjourney produces a lot of unusable images, such as multiple copies of the character reference and blending together of different characters. But it is getting easier all the time. You no longer need to use Discord, complicated setting features, or seed files, for example. You can now drag and drop character references, scene references, and style reference files. Midjourney even added a new tool to edit the portions of an image you don’t like or want to change.

As you can see in the Supplemental Materials provided on the journal website and sample comic panels in Figures 8 and 9, these three graphic novel chapters come with some flaws, such as inconsistent clothing and hair color. We often had to create 20 to 30 different Midjourney images to approximate the characters, scenes, and gestures we wanted and then select those that worked best. We then assembled the images into comic panels using PowerPoint, drew inspiration from various portions of the ChatGPT-generated scripts, and, finally, rewrote most of the text for the comic. The result is a reasonably coherent narrative of the entrepreneurial journey that illustrates various entrepreneurship concepts.

Figure 8.

Representative graphic novel comic panel from the start of Chapter 1.

Figure 9.

Representative graphic novel comic panel from the start of Chapter 2.

The RISE 360 authoring tool was used to create interactive learning objects from these comic book panels and PowerPoint slides, and Articulate Storyline was utilized to create SCORM-compliant content objects that were imported and deployed in the university’s LMS (D2L Brightspace). This student version uses locked navigation features that force students to successfully complete each interactive object, answer all knowledge checks, and pass all quizzes in order to progress to the next lesson or week. We also published an unlocked public version of the interactive content as a web 3.0 online Open Educational Resource (OER) at https://entrepreneurinstitute.ca/building-an-entrepreneurial-mindset-personal-focus/ (Gedeon, 2024).

We wanted our main character (Violet) to make a few of the most common mistakes that we often see our students making. For example, she starts by designing a product for herself rather than interviewing potential customers; she tries to be creative without using any creativity-enhancing techniques; and she uses negative self-talk when things do not go as well as she wants.

We used the “Hot Spot” interactivity features of Articulate Storyline to highlight, identify, and describe the various entrepreneurial concepts being illustrated in the graphic novel. For example, when Violet is trying to force herself to be creative and come up with a great idea in the second panel, our interactive Hot Spot feature tells students that creativity doesn’t happen on its own and that we’ll be running in-class creativity workshops. Creativity is a skill that must be learned and practiced (just like cutting hair or playing the guitar). In the Supplemental Materials on the journal website, we show the Hot Spot comments for the first two chapters directly on the comic panels, as the students would see them. In the third chapter, to reduce space, we show only the location of the Hot Spots in the panels and include the comments themselves in Appendix B at the end of this article.

Between chapters, we pause the action to ask students a series of multiple-response questions to help reinforce the concepts being illustrated by the graphic novel. For example, after the first chapter, we highlight that Violet is demonstrating a planning-based mindset because “She starts by building a final product before she has done any customer discovery or prototyping”; “She starts with an end goal in mind (the App) and then starts to figure out how to do it by learning how to code (controlling the Means)”; and “She is not starting with the end-user in mind. She’s starting with herself in mind.”

At the end of the second chapter, we ask students what advice they might give to Violet and use this opportunity to reinforce certain key themes such as “Don't be discouraged. It is overcoming challenges like this, or pivoting, that causes you to grow your resiliency” and “Violet is using a cognitive distortion called all-or-nothing thinking or catastrophizing. It’s simply not true that ‘nobody liked my idea’. That’s a bit of an exaggeration. Only one person said he would definitely not use it (and that’s a good thing to discover).”

We include the design artifact in the Supplemental Material section on the journal website to demonstrate the current state-of-the-art of what is possible for using AI to create graphic novels by relatively untrained professors. The story could arguably be improved, but, as a prototype, we think it does a decent job of demonstrating key concepts. We also provide a step-by-step how to guide for how we created the graphic novel in Appendix A.

What are your personal experiences with the learning innovation? What worked well and what did not work well?

We have enjoyed the personal learning experience. With all the hype over AI, we wanted to discover for ourselves what is possible and what is not. For one thing, we are not concerned that AI will take over the world in the near term. It can be extremely frustrating to get Midjourney to do anything complicated (such as using multiple characters or illustrating abstract concepts), and you normally need to create dozens of images to find one that works out well enough. The narrative stories created by ChatGPT were, for the most part, weak and simplistic, and asking it to weave entrepreneurial concepts into an overarching story arc was not successful. We had to write most of the story ourselves.

It does not appear to us that AI will replace human imagination, creativity, and expertise anytime soon. However, it is a great tool that enables us to create something we could never have attempted on our own. The ability to create images like Figures 1–7 significantly enhanced the graphic appeal of our slides and educational content and was worth the time and effort.

We solicited informal in-class feedback from our students on the learning innovation. This was done through casual conversations before and after class as well as using Socratic methods during class while showing PowerPoint slides in a 500-seat lecture hall. We did not seek Research Ethics Board approval for a more rigorous analysis.

Around 10–30% of students recognize Figures 1–7, although fewer than 10% volunteered to name the specific entrepreneurial concept associated with an image. Approximately the same percent of students recognized images or characters from the graphic novel. Around 10% of students could name the main characters in the graphic novel and/or describe some of the plot/storyline.

In general, the students don’t appear to notice the visual inconsistencies, even when they are pointed out to them in class. They easily recognize Violet and Bobby as the main characters despite slight changes in their clothing, hairstyle, or features from panel to panel. This demonstrates that character consistency is good enough, notwithstanding current limitations.

The most common feedback was that the entire Articulate Storyline course content is so dense with new ideas, videos, graphics, knowledge checks, and interactive learning objects that these specific images and the graphic novel itself simply blended into the overall learning experience. Each graphic novel chapter was only one lesson out of ten or more lessons in that week’s module. Thus, students saw these 3 lessons out of over 30 lessons in the first 3 weeks as merely an entertaining break in the course content and not something to be remembered or studied in their own right. Several students said they mostly ignore anything that does not directly tie to their grade, and there were no weekly quiz questions specifically related to the graphic novel.

As a result, we do not believe it would be possible to detect any measurable change in student learning outcomes as a direct result of this learning innovation, as distinguished from any change that might be due to the entire course experience. The pilot of this innovation is simply too small of a component of the overall course to make a discernible difference in overall learning outcomes. If we were to undertake a formal research project into the effectiveness of graphic novels, we would want to expand the story to at least eleven full chapters (one for each week’s module), tie the content to the weekly quizzes, and perhaps even highlight the importance of the graphic novel and make it worth at least 10% of the overall course grade to get students’ attention.

Our ad hoc experience is that students appreciate their professors going above and beyond traditional teaching methods. They seem impressed that we can do more than write academic books and give lectures. Some say that they have been inspired to learn more about AI, graphics, prototyping tools, and embracing technology in general. They appear more engaged. It makes us seem a little bit “cooler” in their eyes, and that’s rewarding.

Our initial plan was to introduce several different main characters in separate origin stories, then have them meet, form a team, and work on a startup together. However, at present, we don’t think it is feasible to have more than two characters appear together in a single image without extensive manual editing in software like Photoshop. This will likely improve in the near future.

How can the Learning Innovation Be Applied in Other Contexts?

We published the interactive content as a web 3.0 Open Educational Resource (Gedeon, 2024). Educators may add these free educational assets to their own courses by linking to the online OER or requesting from the authors a SCORM-compliant content object that they can import into their course’s LMS such as Moodle, Blackboard, or D2L Brightspace (locked or unlocked and with or without quizzes).

In addition to these free resources, we want to inspire other professors to create their own AI-generated images and graphic novels. We hope to demonstrate that the learning curve is not too high and getting easier. There are limitations to the current technology, but we believe the outcome is useful and represents a positive advancement in learning innovation that is accessible to all entrepreneurship educators.

Supplemental Material

Supplemental Material - Using AI-Enabled Images and Comic Narrative to Enhance Entrepreneurship Education

Supplemental Material for Using AI-Enabled Images and Comic Narrative to Enhance Entrepreneurship Education by Steven Gedeon and Florian Huber in Entrepreneurship Education and Pedagogy

Supplemental Material

Supplemental Material - Using AI-Enabled Images and Comic Narrative to Enhance Entrepreneurship Education

Supplemental Material for Using AI-Enabled Images and Comic Narrative to Enhance Entrepreneurship Education by Steven Gedeon and Florian Huber in Entrepreneurship Education and Pedagogy

Supplemental Material

Supplemental Material - Using AI-Enabled Images and Comic Narrative to Enhance Entrepreneurship Education

Supplemental Material for Using AI-Enabled Images and Comic Narrative to Enhance Entrepreneurship Education by Steven Gedeon and Florian Huber in Entrepreneurship Education and Pedagogy