They are called DALL-E 2, Midjourney, Craiyon or Stable Diffusion – and they are fascinating examples of what is possible today with artificial intelligence (AI) or machine learning (ML). You enter a short text in English, called a prompt, into these software systems – and after a few seconds or minutes they spit out the appropriate images.
It depends on the input
The result is often astonishing: Photorealistic graphics of non-existent landscapes, oil portraits that would not even have been created with a lot of imagination or just crazy combinations of motifs that shouldn’t really go together. How good the systems are is shown by the fact that some observers are already speculating about the end of art.
But as entertaining – and almost addictive – as the AI-based text-image generators are, they are not easy to use. Because the prompt has to be such that the AI ”understands” it and then really creates the right images. As a result, there are now user-managed databases that can be used for inspiration. They then execute the entered prompts as well as various other configuration features. Is “dog that looks like a giraffe oil painting” better than “giraffe dog oil painting”? What does the generator understand how?
AI helps AI paint pictures
So it would be good if you could let an AI help you to find the right prompts, which then also saves a lot of computing and waiting time. In fact, such systems already exist. The start-up Phraser has developed software that can be accessed via the web and already contains adaptations for various text-image generators – currently DALL-E 2, Midjourney, Stable Diffusion, Disco Diffusion and Craiyon. To create a prompt, you click through a simple menu system.
You can choose from different types of art, such as painting, photography or 3D rendering. Then you enter a first descriptive sentence, for which Phraser also gives examples. Conveniently, examples from images that have already been generated appear parallel to this on the right, so that you can adapt your prompt yourself. Finally, the style, coloring, textures, resolution, emotions to be evoked by the image, and even the era of an image can be selected. So far, however, only users with so-called API access can check how well Phraser actually works – without this, the system does not produce the generated prompt because it can generate images itself via stable diffusion.
To home page