Let's explore how AI can create amazing images, and discuss the important questions about what's real and what's not.
A Tech-Savvy Tuesday!
We've talked about how Artificial Intelligence can understand and use words. Today, let’s explore one of its most talked-about and visually stunning capabilities: creating images from scratch. You may have seen incredible, highly realistic, or wonderfully imaginative pictures online and heard they were "made by AI."
This new technology opens up a new frontier for creativity, but it also brings up important questions about what's real in a digital world. As one of our readers with a background in journalism thoughtfully noted, it can make it hard to "know what's real and what's not real". Let’s take a clear-eyed look at what this technology is, how you can think about using it, and how to be a savvy observer of this new art form.
If you finish this article and want to learn more about AI or try it out yourself, send me a message at Steve@gestalt.com or message me here:
AI Image Generation: Painting with Words
How Does It Work?
At its heart, AI image generation is a process where you give the computer a written description called a "prompt", and the AI creates a brand-new image based on your words.
Think of it like describing a scene to an incredibly fast and imaginative artist who has studied billions of images and their text descriptions from all over the internet. By analyzing this vast library, the AI has learned the relationships between words (like "fluffy," "cat," "sunny") and visual concepts. When you provide a prompt, it uses this knowledge to generate a unique picture that matches your description. These images can be styled to look like anything from a realistic photograph to a watercolor painting or a pencil sketch.
Beyond Creating: Transforming Your Own Photos
This technology isn't just for making images out of thin air. Many tools also allow you to upload one of your own photos and ask the AI to reimagine it in a different style. For example, you could take a simple photo of your cat and give it the prompt, "turn this photo into a vibrant painting in the style of Vincent van Gogh." The AI would then use your photo as a guide to create a completely new, art-inspired version.
(This is my beautiful, grumpy cat Banksy. Or as my Mom calls him— her “grandkitten”.)
Examples to Spark Your Imagination
The magic is in the description. The more detailed your "prompt," the closer the AI can get to the image in your mind. Here are a few examples of prompts and the kinds of images they could create:
For a Realistic Scene:
Prompt:"A photorealistic, sun-drenched image of a classic red barn sitting next to a field of golden wheat, with a bright blue sky and fluffy white clouds overhead. The style should be warm and nostalgic, like a photograph from the 1970s."
For a Stylistic Painting:
Prompt:"A vibrant watercolor painting of a Parisian café scene at dusk. A small, round table with two cups of coffee sits in the foreground, with the warm glow of streetlights reflecting on the wet cobblestone street. The Eiffel Tower is softly visible in the background."
For Something Whimsical:
Prompt:"An adorable, fluffy Corgi wearing a tiny astronaut helmet, floating happily in space with the Earth visible behind it. The style should be charming and gentle, like an illustration from a children's book."
This new technology has sparked a lively and important debate: can something created by an AI truly be considered "art"? It’s a fascinating question with no easy answer. On one hand, some argue that AI is simply a powerful new tool, much like the camera was in its day. They believe the artistry comes from the human who conceives the idea, carefully crafts the descriptive prompt, and curates the final image. As one of our artistically-minded readers noted after seeing an AI exhibit at a museum, it can be "amazingly beautiful" and "fascinating", suggesting the result itself can certainly have artistic merit.
On the other hand, many feel that true art requires a human touch and a lived experience behind it. A reader with a background in journalism and photography raised this thoughtful point, distinguishing between "making a picture" and "taking a picture," emphasizing that a machine can't "feel" what it's creating. Ultimately, it may be a new category of creation altogether, challenging our definitions in the same way photography challenged painting over a century ago.
What do you think?
Copyright and Ownership Concerns
Beyond the philosophical questions, there are also very practical concerns about copyright and ownership that are currently being debated by lawyers, artists, and lawmakers. A central question is: if you create an image using AI, who actually owns it? Is it you, the person who wrote the prompt? Is it the company that built the AI? Or can it even be copyrighted at all?
Another significant concern revolves around how these AI systems were trained. To learn, they analyzed billions of images from across the internet, many of which were copyrighted works by human artists. There is a serious, ongoing discussion about whether it was fair to use that art as training material without the creators' permission or compensation. The rules for this new frontier are still being written in real-time, and for now, it's wise to understand that using AI-generated images, especially for any commercial purpose, exists in a legally gray area.
The Important Conversation: Reality, Trust, and "Deepfakes"
With this amazing creative power comes a serious responsibility. Because these AI-generated images can look so real, it is becoming more difficult to tell the difference between a real photograph and a computer-generated one. This is where we must all become more discerning viewers.
You may have heard the term "deepfake," which refers to realistic but fake images or videos of people (often public figures) created using AI. This technology can be used for harmless fun, but it can also be used to spread misinformation or create malicious content.
The goal is not to be fearful, but to be aware and critical. When you see an extraordinary or surprising image online, especially one involving a public figure, it’s a good habit to pause and ask yourself: Where is this image coming from? Is it from a reputable news source, or from an account known for sharing memes and entertainment? Being a savvy digital citizen means approaching online images with a healthy dose of skepticism and valuing trusted sources.
AI image generation is a remarkable new tool for creativity and imagination. And like any powerful tool, it requires us to use it—and view its creations—thoughtfully and responsibly.
Want to try this out, but don’t know where to get started? Send me an email (or reply to this one) and I’ll help you out!
Steve@Gestalt.com
Quick Tech Tip
When writing a "prompt" for an AI image generator, try to be as descriptive as possible! Don't just say "a house." Instead, try: "a cozy, stone cottage with a thatched roof, smoke curling from the chimney, surrounded by a colorful English garden in the morning light." The more details you provide about the subject, the style, the lighting, and the mood, the closer the AI will get to creating the image in your mind.
Tech Term Demystified: "Prompt" (for AI)
In the world of AI, a "Prompt" is simply the instruction, question, or description you give to the AI to get it started. For a conversational AI like ChatGPT, your prompt is the question you ask. For an AI image generator, your prompt is your written description of the image you want it to create. A good, clear, and detailed prompt is the most important ingredient for getting a great result from any AI tool.
Good News Byte
While some worry about AI replacing artists, many creators are actually embracing it as a powerful new brainstorming partner. Authors are using AI image generators to create "concept art" for the characters and settings in their books, helping them visualize their world more clearly. Architects can generate quick mock-ups of building designs to explore different ideas, and designers can create inspirational "mood boards" in seconds. It's becoming a valuable tool for jump-starting the creative process.
Did You Know?
The core ideas that make modern AI image generation possible have been developing for decades! Early pioneers of "computer art" in the 1960s and 70s wrote complex lines of code to get computers to create simple geometric patterns and text-based art. Today's tools have made that creative process much more accessible, allowing anyone to create complex images using plain, natural language instead of code.
Your Turn to Be the Artist!
You don't need an AI to do this exercise. Think of a favorite memory or an imagined scene. This week, try writing down a short, descriptive paragraph of that scene as if you were describing it to an artist. What are the key objects? What is the lighting like? What is the mood? It's a fun way to practice the art of descriptive language!
A Little Bit of Fun
I asked an AI to create a picture of my perfectly clean and organized desk. It created a beautiful image of an antique mahogany desk with a single, perfectly placed quill pen. I think the AI is trying to politely tell me something.
If you thought this was interesting, your friends might too!
