Real-pedia

The training data for AI large language models mixes up Wikipedia articles with ripped-off text from random authors, toxic trolls off social media, and secret prompts from commercial sponsors. Most of this excludes real knowledge of real places, from the girls excluded from education in Afghanistan, to the women who teach their own families all over the world. Your goal is to make an alternative wiki architecture, where the benchmark of knowledge is not who has the most money or followers, but commonsense human knowledge from real places. You’ll need to think about what motivates people to spend their time this way, what kind of devices they have access to, and what knowledge would get excluded or be unreliable. Your prototype might be a small start, but the AI of the future depends on a better way of understanding what is meaningful.