Generate images with AI

To generate images with AI you describe in words what you want (that's called a prompt) and a tool turns it into an image in a few seconds. You can do it from general-purpose assistants like ChatGPT or Gemini, with specialised tools like Midjourney or even on your own computer with open models. Many also let you start from a photo of your own and ask for changes to it.

Under the hood, the model has learned from millions of captioned images which shapes and colours go with which words, and from there it "paints" something new. What tends to frustrate beginners: text inside the image sometimes comes out with strange letters, hands and small details go wrong, and it's hard to get the same product or character to look the same across several images. Asking for a small change can change the whole image. It helps to describe the style, the framing, the lighting and what you're going to use it for. And a professional logo needs more than a pretty picture: it has to be editable and look good at small sizes.

That's exactly what I set out to test in "My business's images, made with artificial intelligence": logo, product photos and posts, which tools I tried, what came out right first time and what obviously looks like AI. I also go through the legal side there in practical terms, because generating the image is one thing and being able to use it in your business is another.