Human-labeled data currently looks like the quickest path to making robots that are useful enough to have economic value beyond settings and tasks that are expl

Human-labeled data currently looks like the quickest path to making robots that are useful enough to have economic value beyond settings and tasks that are explicitly designed for robots. This has drawn a lot of corporate and academic research activity away from solving the harder core problems, like exploration, that are critical for developing fully autonomous intelligent agents.