This year, the program received more than 11,000 submissions for just 17 spots in its third class. The Impact application consists of 70 questions, what Mitchell describes as a “thesis-like defense” of why the writer’s particular project deserves investment. To manage the overwhelming volume, Impact uses machine learning to sift through the giant pool of applications and identify new voices. The AI is part device, part catalyst: The technology searches for diverse applicants–Mitchell explains they look for people who have overcome challenges in their careers or lives–with the goal of shaking up the historically homogenous film industry. – Wired