Karen Yeung mentioned the fact that even within the most complex and evoluted machine-learning system, a human decisional capacity is still needed to curate the first outputs, and incrementally increase the pertinence of the AI work. But will this be true forever? Or will we live to see a day when news will be written by algorithms? Le Monde already uses automated algorithmic systems to write basic/simple articles such as election results. And if it happens to be a wider phenomenon extending to other professional fields, how can we re-think the place that the human would take within this situation?