No doubt that AI has developed tremendously in the past months, but still when asked to generate a specific, culturally significant visual image, it appears that text-to-image generative AI models frequently create images that reflect biased and stereotypical characteristics associated with gender, skin color, occupations, nationalities, and other attributes. As for an example, I present an image generated through Midjourney, for a prompt asked to imagine “A typical Latvian family posing in front of her country side house with a forest view in the back, captured in a photorealistic style”. As it can be seen, many characteristics and items visible in the image are misleading and not projecting a truthful description of a typical Latvian family.