Washington Post [read free]: “…Stability AI, maker of the popular image generator Stable Diffusion XL, told The Washington Post it had made a significant investment in reducing bias in its latest model, which was released in July. But these efforts haven’t stopped it from defaulting to cartoonish tropes. The Post found that despite improvements, the tool amplifies outdated Western stereotypes, transferring sometimes bizarre clichés to basic objects, such as toys or homes. “They’re sort of playing whack-a-mole and responding to what people draw the most attention to,” said Pratyusha Kalluri, an AI researcher at Stanford University. Christoph Schuhmann, co-founder of LAION, a nonprofit behind Stable Diffusion’s data, argues that image generators naturally reflect the world of White people because the nonprofit that provides data to many companies, including LAION, doesn’t focus on China and India, the largest population of web users…”