v21’s Twitter Archive—№ 78,059

you know how Uber's strategy for dealing with the fact that their business model relied on breaking labour law was to race to get big enough to change the law before they got shut down? feels like companies training AI models on copyrighted data are trying the same strategy