YouTube: "identifying copyrighted material can't be an automated process." Startup disagrees. | Hacker News Reader