This pictorial investigates how overlooked personal habits and machine-learning biases in photo-imaging practices can be identified through the image-generative capabilities of a GAN algorithm. Incorporating a designerly way of reflecting is key to noticing such habits and biases. These are therefore revealed both by noticing what the algorithm is capable of, yet also the relationship between humans and the world which this technology mediates. Practices of photo-taking and photo-looking facilitate this process of reflection. Simultaneously, the technology provides a window into its functioning and its potential integration into on-site image-taking practices.