Wie Arbeiter die Humanoiden ausbilden, die ihre Jobs übernehmen

From the article

>Carnegie Mellon University (CMU) researchers have developed H2O – Human2HumanOid – a reinforcement learning-based framework that allows a full-sized humanoid robot to be teleoperated by a human in real-time using only an RGB camera. Which begs the question: will manual labor soon be performed remotely?

Also from the article

>This brings us back to the earlier question: Could H2O, or something like it, enable remote labor? CMU’s announcement regarding H2O has generated some thought-provoking discussion on Reddit’s [](https://www.reddit.com/r/Futurology/comments/1bzo3oi/remote_working_might_be_possible_for_manual_jobs/) subreddit.

>Obviously, using teleoperated humanoids means humans get to keep their jobs (a positive). Another plus is employing these humanoids for risky work or search-and-rescue missions in remote and/or human-inaccessible regions.

>However, there are some downsides, the most apparent one being related to labor. As pointed out by Reddit user [lughnasadh](https://www.reddit.com/user/lughnasadh/), “[I]t opens up more avenues for job offshoring in rich countries.” Then there’s the slippery-slope argument: Teleoperated robots now, fully autonomous (read: job-stealing) robots later.

>Interestingly, a [2022 study](https://www.sciencedirect.com/science/article/pii/S2405896322026015) to gauge public acceptance of using teleoperated robots to perform labor jobs found that many people are open to remote work with lower pay than onsite roles. Again, it’s one of those ‘time will tell’ situations we’ve become accustomed to in the swiftly advancing AI/robotics space.

>