We already have “24/7 self-charging robot workers”, they’re called fixed robotic arms and they work amazingly well for repetitive tasks. Humanoid robots are interesting as a novelty, but I’m not sure they have much practical use. Best case is if they’re able to be installed into a factory with less retrofitting than full fixed roboticization, but I have my doubts of that being more cost-effective (or working at all, for that matter).
The advantage of humanoid robots is their versatility. A single robot can be adapted to accomplish many different tasks. I don’t expect these will compete with fixed robotic arms, but rather complement them the way humans do right now.
That’s the supposed benefit, but to do that you’d have to make them more cost-effective than just having humans doing the job, which so far has completely failed to happen. Unless I missed something, there are a whole bunch of breakthroughs that haven’t yet that’d be required for that to work.
UniTree produces humanoid robots for around 12k a pop already. These types of robots can already do lots of useful jobs that humans used to do. It’s not some future thing that requires breakthroughs, it’s happening already. Here’s one example https://peertube.mesnumeriques.fr/w/jNakB3B5dRoeJkShnVRgDk
This tech is rapidly improving, and it’s only going to keep getting better going forward which means robots taking over an increasingly wide range of manual labour. Machine learning is also advancing rapidly right now, and new things are being figured out every day. These types of breakthroughs are happening on monthly basis right now https://github.com/sizhe-li/neural-jacobian-field
We already have “24/7 self-charging robot workers”, they’re called fixed robotic arms and they work amazingly well for repetitive tasks. Humanoid robots are interesting as a novelty, but I’m not sure they have much practical use. Best case is if they’re able to be installed into a factory with less retrofitting than full fixed roboticization, but I have my doubts of that being more cost-effective (or working at all, for that matter).
The advantage of humanoid robots is their versatility. A single robot can be adapted to accomplish many different tasks. I don’t expect these will compete with fixed robotic arms, but rather complement them the way humans do right now.
That’s the supposed benefit, but to do that you’d have to make them more cost-effective than just having humans doing the job, which so far has completely failed to happen. Unless I missed something, there are a whole bunch of breakthroughs that haven’t yet that’d be required for that to work.
UniTree produces humanoid robots for around 12k a pop already. These types of robots can already do lots of useful jobs that humans used to do. It’s not some future thing that requires breakthroughs, it’s happening already. Here’s one example https://peertube.mesnumeriques.fr/w/jNakB3B5dRoeJkShnVRgDk
This tech is rapidly improving, and it’s only going to keep getting better going forward which means robots taking over an increasingly wide range of manual labour. Machine learning is also advancing rapidly right now, and new things are being figured out every day. These types of breakthroughs are happening on monthly basis right now https://github.com/sizhe-li/neural-jacobian-field