Craiyon, an AI image generator

We can’t call Craiyon a real tool yet, in the sense of a professional instrument that can be used to solve critical problems. No, Craiyon is a bit more like a toy, at least for the time being in its rather early stage of development, but Craiyon is a demonstration of what we can expect in the future. Craiyon, is an alternative version of DALL·E mini, by the same developers. It is an AI model that can draw images from any text prompt that you enter. In our banner example we used “a Picasso painting of a teacher with a laptop in his hands”. The Craiyon model is free to use for non-commercial purposes but sponsored by ads. It is trained using the Google PU Research Cloud (TRC). The capabilities of image generation models are impressive, but there are some limitations: the model surprises on the one hand with unexpected creations but seems limited to the common stereotyping and imagery that dominates the sources from which the model is learning. Its artificially created images and especially the faces can look very weird at times. Free at www.craiyon.com.