10-15 Minute Lesson Script
Introduction (2 minutes)
Welcome to Lesson 7! Last week, you learned how to design graphics in Canva using templates and AI tools. This week, we’re taking things to a whole new level: creating images from nothing but text descriptions.
Imagine typing “a cozy coffee shop on a rainy autumn day, watercolor style” and watching an AI create that exact image in seconds. That’s what we’re exploring today with DALL·E and Bing Image Creator.
This technology feels like magic—and honestly, it kind of is. You’re going to describe images in words, and artificial intelligence will paint, draw, or render them for you. No camera needed. No artistic skills required. Just your imagination and clear descriptions.
Here’s what’s exciting: these aren’t just fun toys. You can use AI-generated images for blog headers, social media posts, presentations, product mockups, book covers, and so much more. If you’ve ever needed an image but couldn’t find the right stock photo—or couldn’t afford custom artwork—this lesson will change everything.
By the end of today, you’ll understand how AI image generation works, how to write effective image prompts, and how to create custom visuals for any project.
Let’s create some art!
What Is AI Image Generation? (2 minutes)
AI image generation tools use artificial intelligence to create completely new images based on text descriptions you provide. These aren’t pulled from a database of existing photos—the AI actually creates them from scratch.
How it works (simplified): The AI has been trained on millions of images with descriptions. It learned what “cozy” looks like, what “watercolor style” means, what “autumn” involves, and how coffee shops are typically arranged. When you describe what you want, the AI combines all this learned knowledge to generate something new.
The two main tools we’re covering today:
DALL·E (by OpenAI)
- Made by the same company that created ChatGPT
- Known for creative, artistic results
- Great for illustrations, concepts, and stylized images
- Available through ChatGPT Plus or directly at labs.openai.com
Bing Image Creator (powered by DALL·E)
- Free to use with a Microsoft account
- Also uses DALL·E technology
- Integrated with Bing Chat
- Generates 4 image variations per prompt
- Great starting point before investing in paid tools
Important to know: Both tools have the same underlying technology, but Bing Image Creator is free and gives you plenty of daily credits to experiment with. That’s where we’ll focus most of our attention.
Understanding Image Prompts (3 minutes)
Creating good AI images is all about the prompt—just like everything else you’ve learned in Phase 1. But image prompts have their own special structure.
The anatomy of a great image prompt:
1. Subject – What’s the main focus?
- “A golden retriever”
- “A vintage bicycle”
- “A stack of books”
2. Setting/Context – Where is this happening?
- “…in a sunny garden”
- “…on a cobblestone street”
- “…on a wooden desk by a window”
3. Style – What artistic approach?
- “…watercolor painting”
- “…digital art”
- “…photograph”
- “…pencil sketch”
4. Details – What specific elements matter?
- “…with autumn flowers blooming”
- “…during golden hour sunset”
- “…with morning coffee and reading glasses nearby”
5. Mood/Atmosphere – What feeling should it convey?
- “…peaceful and serene”
- “…energetic and vibrant”
- “…cozy and inviting”
Putting it together:
Basic prompt: “A coffee shop”
Better prompt: “A cozy coffee shop interior, warm lighting, autumn decorations, watercolor painting style, inviting atmosphere”
See how much more specific the second version is? The AI knows exactly what to create.
Another example:
Basic: “A garden”
Better: “A cottage garden overflowing with colorful flowers, stone pathway, morning sunlight filtering through trees, impressionist painting style, peaceful mood”
The more specific you are about what you want, the closer the AI gets to your vision.
Writing Effective Image Prompts (3 minutes)
Let me share some techniques that will dramatically improve your results:
Technique #1: Start with style Beginning your prompt with the artistic style often gives better results.
- “Oil painting of a mountain landscape at sunset”
- “Vintage photograph of a 1950s diner”
- “Minimalist illustration of a coffee cup”
Technique #2: Use specific artistic references Mentioning art movements or famous styles helps the AI understand the aesthetic you want.
- “Art nouveau style poster”
- “Studio Ghibli anime style”
- “Norman Rockwell illustration”
- “Ansel Adams black and white photography”
Technique #3: Control composition Tell the AI how you want the image framed.
- “Close-up of…”
- “Wide angle view of…”
- “Bird’s eye view of…”
- “Portrait orientation” or “Landscape orientation”
Technique #4: Specify lighting Lighting dramatically affects mood.
- “Golden hour lighting”
- “Soft natural window light”
- “Dramatic shadows”
- “Bright and airy”
Technique #5: Add mood and emotion Help the AI understand the feeling.
- “Peaceful and calm”
- “Energetic and vibrant”
- “Mysterious and moody”
- “Warm and inviting”
Example prompt using all techniques:
“Watercolor painting of a cozy home library, floor-to-ceiling bookshelves, comfortable reading chair by window, warm afternoon sunlight streaming in, house cat sleeping on ottoman, peaceful and serene atmosphere, soft colors, intimate close-up view”
This prompt gives the AI everything it needs: style, subject, setting, lighting, details, mood, and composition.
Using Bing Image Creator (3 minutes)
Let’s walk through actually creating images with Bing Image Creator since it’s free and accessible to everyone.
Getting started:
- Go to bing.com/create (or search “Bing Image Creator”)
- Sign in with a Microsoft account (free to create if you don’t have one)
- You’ll see a prompt box that says “Describe the image you want to create”
Creating your first image:
- Type your prompt in the box
- Example: “Watercolor illustration of a garden tea party, vintage china, flowers on table, soft pastel colors, afternoon sunlight, whimsical and cheerful”
- Click “Create”
- Wait 10-30 seconds (processing time varies)
- Review your 4 image options
- Bing creates 4 variations of your prompt
- Each will be slightly different
- Click on any image to see it full-size
- Download your favorite (click the download icon)
Understanding your results:
Sometimes you’ll get exactly what you imagined. Other times, you’ll get something interesting but different. That’s normal! AI image generation involves some randomness.
If the results aren’t quite right:
- Adjust your prompt – Add more specific details
- Try again – Each generation is different
- Refine the style – Be more specific about artistic approach
- Simplify – Sometimes less is more; overly complex prompts can confuse the AI
Daily credits:
Bing gives you “boosts” that speed up generation. When you run out, generation just takes a bit longer. You get more boosts daily, so you can create lots of images for free.
Common Mistakes and How to Fix Them (2 minutes)
Let me save you some frustration by covering common beginner mistakes:
Mistake #1: Being too vague
- ❌ “A nice scene”
- ✅ “A peaceful lakeside cabin at sunset, pine trees, reflections on water, oil painting style”
Mistake #2: Asking for text in images
- AI struggles with text and often gets letters wrong
- If you need text, add it later in Canva or another tool
- ❌ “Book cover with the title ‘My Story’ written on it”
- ✅ “Book cover design with space for title, vintage aesthetic, leather texture” (add text yourself later)
Mistake #3: Too many competing elements
- ❌ “A dog, cat, bird, and hamster at a birthday party in a garden with a rainbow and fireworks”
- ✅ “A golden retriever at a backyard birthday party, balloons and cake on table, sunny afternoon”
Mistake #4: Expecting photorealism for everything
- Some styles work better than others
- Illustrations, paintings, and stylized art often look better than attempts at photorealism
- Start with illustrated styles before attempting realistic photos
Mistake #5: Not iterating
- Your first attempt rarely gives perfect results
- Generate multiple versions
- Adjust your prompt based on what you see
- Think of it as a conversation with the AI
Practical Applications (2 minutes)
Let’s talk about how you’d actually use these AI-generated images:
For blogs and websites:
- Custom featured images that match your exact topic
- Hero images for landing pages
- Illustrations for articles
- Icons and decorative elements
For social media:
- Instagram post backgrounds
- Quote graphics (generate background, add text in Canva)
- Story templates
- Themed content series
For presentations:
- Slide backgrounds
- Concept illustrations
- Visual metaphors for complex ideas
For products:
- Book cover concepts
- Product packaging mockups
- Marketing materials
- Printables and digital downloads
For creative projects:
- Inspiration boards
- Character concepts for stories
- Scene visualization for creative writing
- Gift ideas (print and frame AI art)
Real workflow example:
You’re writing a blog post about morning routines:
- Generate in Bing: “Peaceful morning scene, coffee cup on wooden table, journal and pen, soft morning light through window, watercolor illustration, warm colors”
- Download your favorite version
- Upload to Canva and add your blog title text
- Use as featured image
Total time: 5 minutes. Cost: $0.
DALL·E vs. Bing Image Creator: Which to Use? (1 minute)
Start with Bing Image Creator because:
- It’s completely free
- Gives you 4 variations per prompt
- Powered by the same DALL·E technology
- No credit card required
- Generous daily usage
Consider DALL·E (through ChatGPT Plus) if:
- You’re already paying for ChatGPT Plus
- You want it integrated with ChatGPT conversations
- You need more control and editing features
- You’re using images commercially and want clearer licensing
For most people starting out: Bing Image Creator is perfect. Learn the ropes there, generate hundreds of images for free, and only upgrade if you find yourself needing more.
Wrapping Up (1 minute)
Creating images from text descriptions still feels like science fiction, but it’s real and it’s available to you right now, for free.
The key to success is thinking about image creation the same way you think about prompt crafting for ChatGPT:
- Be specific
- Include style, mood, and details
- Iterate and refine
- Experiment and have fun
This week, your mini challenges will have you generating different types of images and discovering what works best for your needs. By the end of the week, you’ll have a collection of custom images and the skills to create more whenever you need them.
Remember: AI-generated images aren’t meant to replace photographers or artists. They’re tools that expand what’s possible for people who need custom visuals but don’t have the budget or skills to create them otherwise.
Let’s go make some art!
3-5 Key Takeaways
- AI image generation creates brand new images from text descriptions – These aren’t stock photos pulled from a database; the AI actually generates original images based on what you describe, giving you custom visuals without photography or artistic skills.
- Effective image prompts include five key elements – Subject (what), setting (where), style (artistic approach), details (specific elements), and mood (feeling). The more specific your prompt, the closer the result matches your vision.
- Bing Image Creator is the best free starting point – It uses DALL·E technology, creates 4 variations per prompt, has generous daily limits, and costs nothing. Perfect for learning and experimentation before investing in paid tools.
- Iteration is essential for good results – Your first generation rarely gives perfect results. Adjust your prompt based on what you see, generate multiple versions, and refine your descriptions just like you learned with text-based AI.
- AI-generated images work best for illustrations and stylized art – While photorealism is possible, illustrated styles, watercolors, digital art, and paintings typically produce more consistent, professional-looking results. Start with these styles before attempting realistic photography.
2-3 Practical Examples/Case Studies
Example 1: Margaret’s Cookbook Blog Transformation
Background: Margaret, 62, writes a food blog featuring family recipes. She’d been using generic stock photos of food because she wasn’t confident in her food photography skills, and her photos never looked as appetizing as professional food bloggers’ images.
Her challenge: Create unique, appealing visual content that made her blog stand out without learning professional food photography.
Her AI image journey:
Week 1 – Discovery:
- Started with Bing Image Creator
- First prompt (too vague): “A cake”
- Result: Generic, not useful
Week 2 – Getting specific:
- Improved prompt: “Homemade chocolate layer cake on vintage cake stand, afternoon window light, cozy kitchen setting, warm tones, food photography style, appetizing and inviting”
- Result: Beautiful, exactly the aesthetic she wanted
Week 3 – Building a system:
- Created prompt templates for different recipe types
- Recipe header template: “[Food item] on rustic wooden table, natural lighting, herbs and ingredients nearby, food photography style, warm and inviting”
- Generated 2-3 options per recipe, chose the best fit
Her evolved workflow:
- Write recipe post
- Generate 3-4 AI images using her template (5 minutes)
- Choose best image and upload to Canva
- Add recipe title and simple text overlay in Canva (3 minutes)
- Use as featured image
Results:
- Blog traffic increased 65% in 3 months
- Readers commented that her blog “looks so professional now”
- She created a consistent visual brand without expensive equipment
- Time per post decreased (no more struggling with phone camera and lighting)
Her surprise discovery: She started generating seasonal header graphics for holidays:
- “Thanksgiving feast illustration, autumn colors, watercolor painting style, warm and festive”
- “Christmas cookie decorating scene, cozy kitchen, holiday atmosphere, vintage illustration”
These seasonal images got shared on Pinterest hundreds of times.
Key lesson: AI image generation solved Margaret’s problem (inconsistent visuals) without requiring her to learn a new technical skill (photography). She focused on what she’s good at (recipes) while AI handled what she struggled with (appealing food visuals).
Example 2: Robert’s Small Business Marketing
Background: Robert, 56, owns a lawn care company. His social media posts used either his own phone photos (often unflattering) or generic stock photos that didn’t represent his actual work style.
His challenge: Create professional-looking social media content consistently without hiring a designer or photographer.
His strategic approach:
First, he identified his brand aesthetic:
- Wanted to convey: reliability, quality, connection to nature
- Style: Clean, professional, not too corporate
- Colors: Greens, earth tones, natural light
Then created a prompt library:
Template 1 – Seasonal service posts: “Professional illustration of [seasonal task, e.g., ‘autumn leaf cleanup’], beautiful residential yard, satisfied homeowner in background, warm natural lighting, digital art style, clean and professional”
Template 2 – Before/After concepts: “Split view illustration showing overgrown lawn on left transforming into beautifully maintained lawn on right, suburban house, clear improvement, professional photography style”
Template 3 – Equipment and expertise: “Professional lawn care equipment in action, well-maintained yard, expert at work, action shot, bright clear day, digital illustration, competent and trustworthy”
His content calendar:
- Monday: Weekend recap (AI-generated scene + description of work done)
- Wednesday: Seasonal tip (AI-generated helpful illustration)
- Friday: Before/after showcase (AI concept + real results description)
Time investment:
- Old way: 2+ hours per week struggling with photos and design
- New way: 30 minutes on Sunday generating all week’s images, 5 minutes per post adding text in Canva
Results:
- Social media engagement tripled
- Booked 8 new clients who specifically mentioned his “professional online presence”
- Competitor asked him who his marketing person was (it was just Robert and AI)
- Actually enjoyed social media instead of dreading it
Key lesson: Small businesses can compete visually with larger companies by using AI to create professional marketing materials. Robert didn’t need to fake having a marketing department—he just needed smart tools that amplified his own efforts.
Example 3: Diana’s Children’s Book Journey
Background: Diana, 59, had written a children’s story for her grandchildren but couldn’t afford to hire an illustrator ($2,000-5,000 minimum).
Her challenge: Create illustrated children’s book pages that were consistent in style and brought her story to life.
Her approach:
Phase 1 – Style discovery (1 week): Experimented with different illustration styles:
- “Watercolor children’s book illustration”
- “Vintage storybook art style”
- “Modern children’s book illustration, bright colors”
- “Whimsical cartoon style”
Discovered she loved watercolor style for her gentle story about a garden.
Phase 2 – Character consistency (2 weeks): The hardest part: keeping her main character (a little rabbit) looking the same across images.
Her solution: Created a detailed character description she used in every prompt: “Small brown rabbit with white cottontail, wearing a blue vest, friendly expression, [doing action], watercolor children’s book illustration, soft colors, whimsical style”
Generated 10+ versions until she found a character design she loved, then used that exact prompt for all future images.
Phase 3 – Scene creation (3 weeks): Created 15 scene illustrations for her book:
- Opening: “Small brown rabbit with white cottontail and blue vest, standing at garden gate, looking curious, morning light, watercolor children’s book illustration”
- Middle: “Small brown rabbit with blue vest, discovering colorful flowers, surprised and delighted expression, garden setting, watercolor style”
- Ending: [similar detailed prompt for concluding scene]
Her publishing path:
- Generated all illustrations using Bing Image Creator (free)
- Edited and arranged in Canva
- Self-published through Amazon KDP
- Total cost: $0 for illustrations (would have been $2,500+ for professional illustrator)
Results:
- Published her book and gave copies to all 7 grandchildren
- Sold 150+ copies to friends and community members
- Started working on a second book
- Most importantly: Her dream of being a published author came true
Challenges she faced:
- Consistency took multiple attempts
- Some scenes needed 10+ generations to get right
- Had to accept “good enough” instead of perfect
- Learned that detailed prompts were essential
Key lesson: AI image generation can make previously impossible dreams achievable. Diana’s story would have remained unillustrated without AI, because a $2,500 illustration budget wasn’t realistic. AI didn’t just save her money—it made her dream possible.
1 Hands-On Activity
Activity: “Build Your AI Image Portfolio”
Time needed: 45-60 minutes
Objective: Generate multiple images across different styles and use cases, learning what works and building confidence with AI image creation.
What you’ll need:
- Computer or tablet with internet
- Microsoft account (free to create)
- Notebook or document for tracking your prompts and results
Part 1: Set Up Bing Image Creator (5 minutes)
- Go to bing.com/create
- Sign in with your Microsoft account (create one if needed—it’s free)
- Familiarize yourself with the interface:
- Prompt box at the top
- “Create” button
- Your image history below
- Note your daily “boosts” (shown in the corner)—these make generation faster
You’re ready to create!
Part 2: The Style Exploration Challenge (15 minutes)
Goal: Generate the SAME subject in 4 different artistic styles to understand how style affects results.
Choose a simple subject: A coffee cup, a flower, a book, a bicycle, or a cabin
Generate 4 versions, each with a different style:
Version 1 – Watercolor: “[Your subject], watercolor painting style, soft colors, delicate brush strokes, artistic”
Version 2 – Digital illustration: “[Your subject], modern digital illustration, clean lines, vibrant colors, professional”
Version 3 – Vintage photograph: “[Your subject], vintage 1970s photograph, film grain, muted colors, nostalgic”
Version 4 – Pencil sketch: “[Your subject], detailed pencil sketch, shading, artistic drawing, black and white”
For each generation:
- Write down your exact prompt
- Note which of the 4 variations you like best
- Download your favorite
Reflection: Which style looks most professional? Which matches your aesthetic? Which surprised you?
Part 3: The Real-World Application Challenge (20 minutes)
Goal: Create 3 images you could actually use for real projects.
Think about YOUR needs: What do you actually need images for?
- Blog posts?
- Social media?
- Presentations?
- Personal projects?
Create 3 different images:
Image 1 – Blog/Article header:
Think of a topic you write or talk about. Create a header image for it.
Template prompt: “[Topic visualization], [setting/context], [artistic style], [lighting], [mood], professional quality, suitable for blog header”
Example: “Cozy home office with laptop and coffee, morning sunlight through window, digital illustration style, warm and inviting, professional quality”
Image 2 – Social media post background:
Create a background you could use for a quote, tip, or announcement.
Template prompt: “[Simple scene or pattern], [color palette], [style], clean composition with space for text overlay, [mood]”
Example: “Abstract watercolor background, soft pastel blues and greens, flowing organic shapes, peaceful and calm, with negative space for text”
Image 3 – Concept illustration:
Choose something abstract or conceptual to visualize.
Template prompt: “Visual metaphor for [concept], [artistic approach], [style], clear and symbolic, [mood]”
Example: “Visual metaphor for personal growth, small plant growing into strong tree, sunrise background, hopeful and inspiring, watercolor illustration style”
For each image:
- Write the prompt
- Generate and review the 4 options
- Download the best one
- Note what you’d use it for
Part 4: The Iteration Challenge (15 minutes)
Goal: Learn to refine prompts by starting vague and getting progressively more specific.
Choose ONE topic and generate it 3 times, improving each time:
Attempt 1 – Vague prompt: Start with something basic like “a garden” or “a kitchen” or “a workspace”
Generate it and see what you get.
Attempt 2 – Add specific details: Take your first prompt and add:
- Specific time of day
- Style preference
- Mood/atmosphere
- 2-3 specific elements
Example evolution: “A garden” → “A cottage garden in spring, morning sunlight, colorful flowers blooming, stone pathway, peaceful atmosphere, watercolor painting style”
Attempt 3 – Add composition and refined details: Take your second prompt and add:
- Composition direction (close-up, wide angle, etc.)
- Lighting specifics
- Color palette guidance
- Refined mood descriptors
Example evolution: “A cottage garden in spring, morning sunlight, colorful flowers blooming, stone pathway, peaceful atmosphere, watercolor painting style” → “Close-up view of spring cottage garden, golden morning light filtering through trees, pink and purple flowers in bloom, weathered stone pathway, cozy and inviting atmosphere, soft pastel watercolor illustration, intimate perspective”
Compare all three versions:
- How did the results change?
- Which prompt gave the best result?
- What specific details made the biggest difference?
Part 5: Create Your Prompt Library (5 minutes)
Based on everything you’ve generated:
Document your best prompts:
- Best overall prompt:
- [Write the complete prompt]
- Why it worked: [Your notes]
- Best style descriptor:
- What style wording gave you the results you liked best?
- Best mood/atmosphere words:
- Which emotional descriptors worked well?
- Templates you’ll reuse: Create 2-3 prompt templates with [brackets] for things you’ll change: Example template: “[Subject] in [setting], [time of day], [artistic style], [mood], [specific details], professional quality”
Save these! You’ll use them again and again.
Part 6: Reflection and Planning (5 minutes)
Look at all the images you created today (should be 10-15 images):
- Rate your confidence:
- Before: How confident were you that you could create custom images? (1-10)
- After: How confident are you now? (1-10)
- Identify your favorite:
- Which single image are you most proud of?
- What made it work?
- Plan your next creations:
- What 3 images will you create next week for real projects?
- Write down the prompts you’ll try
- Note your learnings:
- What surprised you about AI image generation?
- What was harder than expected?
- What was easier than expected?
- Which style do you naturally gravitate toward?
Bonus Challenge:
Take your best AI-generated image and bring it into Canva (from Lesson 6):
- Add text overlay
- Adjust colors if needed
- Create a complete social media post or blog header
- Save it for actual use
Congratulations! You’ve just created a portfolio of custom images using only text descriptions. That’s incredible! 🎨
3 Quiz Questions with Answers
Question 1
What makes a good AI image prompt effective?
A) Using as many words as possible to describe everything
B) Including five key elements: subject, setting, style, details, and mood
C) Using technical art terminology that only professionals understand
D) Keeping it as simple and vague as possible to let the AI be creative
Answer: B – Including five key elements: subject, setting, style, details, and mood
Explanation: Effective image prompts are specific but structured. They include what you want (subject), where it is (setting), how it should look (style), what specific elements to include (details), and what feeling it should convey (mood). This gives the AI clear direction while providing enough information to create what you envision. Too vague (“a garden”) gives unpredictable results, while too complex or technical can confuse the AI. The sweet spot is specific, clear descriptions using these five elements.
Question 2
Why is Bing Image Creator recommended for beginners over paid alternatives like DALL·E?
A) It produces better quality images than paid options
B) It’s completely free, uses the same technology as DALL·E, and has generous daily limits
C) Paid AI image tools don’t work as well
D) You must start with Bing before you’re allowed to use other tools
Answer: B – It’s completely free, uses the same technology as DALL·E, and has generous daily limits
Explanation: Bing Image Creator is powered by the same DALL·E technology that paid services use, so the quality is comparable. The advantage for beginners is that it’s free, requires only a Microsoft account, and gives you plenty of daily generations to learn and experiment. You can create hundreds of images without spending money, which is perfect for building skills and discovering what works. Once you know how to write effective prompts and understand your needs, you can decide if paying for premium features is worthwhile. Starting free lets you learn without financial risk.
Question 3
What should you do if your first AI-generated image doesn’t match what you envisioned?
A) Give up—the AI obviously can’t create what you want
B) Keep using the exact same prompt and hope it works eventually
C) Adjust your prompt with more specific details and try again, just like iteration with text AI
D) Switch to a completely different subject
Answer: C – Adjust your prompt with more specific details and try again, just like iteration with text AI
Explanation: AI image generation works on the same iteration principle as the text AI you learned in Phase 1. Your first attempt rarely gives perfect results—that’s normal and expected. The smart approach is to analyze what the AI created, identify what worked and what didn’t, then refine your prompt with more specific details about style, lighting, composition, or mood. Each generation teaches you how the AI interprets your words, helping you improve your prompts. This iterative process is how experienced users get great results, not by writing perfect prompts on the first try.