Keeping It Real: Maintaining the Personal Touch When Building Courses With AI
As a lifelong Garfield aficionado, I am wondering if using AI in course creation is like making lasagna.
When you make a lasagna, you have loads of choices about how much of it you are going to do yourself. Will you roll the pasta yourself or use lasagna sheets from a packet? Will you make both sauces yourself, or skip the added effort of making the bechamel sauce? Maybe you’ve even been known to just buy a pre-made lasagna, and add a couple of finishing ingredients to it.
Either way, technology will have been used for some of those parts. Whether or not you made the pasta sheets yourself, you almost certainly didn’t harvest and mill the durum wheat. Whether it’s beef mince or vegetarian mince, it was probably minced before you bought it.
As AI becomes embedded in our work, the question is less whether we use technological shortcuts and more how we use them. But if everyone is cooking with the same ingredients, following the same recipe, and using the same equipment, how do we ensure that our work retains its distinct flavour and impact? Let’s think about how to keep it real.
Think about your style
How do you like your lasagnas? Maybe you like them to be rich and indulgent, with extra layers of different types of cheese. Maybe you prefer spicy and bold, with plenty of chili, pepper, or spiced meats. Or perhaps you like something rustic and hearty, with chunky vegetables and thicker-cut meat.
Likewise, which words should describe the feel of your course? The tone could range from informal to formal, or from conversational to academic. It might contain more dynamic, attention-grabbing text and imagery, or it could instead be a calm, methodical approach. Once you have a better idea of the flavour you’re looking for, you can instruct the AI to keep that style in mind. Example prompts:
- Use a formal, academic tone with precise language, avoiding contractions, and citing evidence-based practices where appropriate.
- Keep the tone professional but approachable, using plain language, real-world examples, and an encouraging style that makes employees feel engaged.
Avoiding the dead giveaways
If you enjoy a home-made lasagna, then you would be able to identify a factory-made lasagna masquerading as one that’s been artfully prepared in the kitchen. Maybe it’s too salty or too sweet, or perhaps the shape is just a little too regular. Or maybe it was made in the kitchen, but you have your suspicions about where the shortcuts were taken.
Likewise, we’re all becoming better at spotting text that looks like it's come from a large language model. You might spot things like:
- Overuse of em-dashes—like the ones either side of these words—within sentences
- A sentence structure that is too perfectly balanced, or used repeatedly (For example, the "it isn't just X, it's Y" structure. AI isn't just handy, it's essential.)
- Lists
- Lists within lists
- Lots of lists
- Emoji in lists 😲
- Overcompensation when directly addressing the reader, to the point that it becomes insincere. Using a phrase like Look at it this way might seem jarring if the rest of the text doesn't feel like it has been written by a human.
I'm sure there are plenty more. If you can think of other AI giveaways, I'd love to hear about them in the comments.
Be transparent
Going back to that home-made lasagna. None of us would appreciate it if we were served a lasagna that had supposedly been cooked from scratch for several hours, but which had actually spent a few minutes in the microwave after its journey from the freezer.
The implication here is to be honest about where you have chosen to use the technology. By modelling ethical and responsible use of AI, you can reassure the learners that you are an honest guide in their learning journey. Acknowledging where AI has been used can also prevent learners from becoming distracted by speculating whether the content is too polished or generic to have been written by a person. It is ok to use technology to take shortcuts, but it’s better to be honest about where you’ve done this.
Keep learner empathy at the forefront
Your lasagna is more likely to be appreciated if you know the tastes of those that you’re catering for.
Good course design begins with knowing who your learners are. What motivates them? What challenges do they face? What language feels natural to them? As you use AI, ensure that it understands the audience, and keep checking that what it generates will resonate with them. The time that you save from using AI can be channeled into keeping content updated to reflect current developments.
Include humans in the course
Humans are storytellers, and we often want to see and hear from other humans. I will probably appreciate a lasagna even more if I know that the San Marzano tomatoes were hand-picked by members of a family who have been growing tomatoes for eight generations. We want to learn from shared wisdom, expertise, and experience. Welcome videos, audio explanations, case studies, anecdotes, and quotes can bring a course to life and make it more memorable and human, so it’s good to get these in amidst the AI-generated content.
What are your thoughts?
How do you balance the use of AI with the human touch? Do you have strong opinions about lasagna making? I’d love to hear your thoughts.
AI disclaimer: For this article, I used AI if I was making a meal of a sentence (so to speak) and wanted a more concise version. I asked it to link a couple of my thoughts to the lasagna analogy, but I rejected the results. The structure of this word-lasagna, and the choice of analogy, were my own. The images are all generative AI.