Continuing the discussion from Leave the birthrate alone:
What are among those many activities?
And among them which could still have productivity significantly increased with AI enabled robotic assistance if cost effective?
Many trades are on the list that human form and adaptability seems to be great for, but if there comes an era of few workers and cheaper AI robotics could they at least enhance human worker ability to get more done?
One possibility might be social or social-like interactions with other humans. We react differently to things that look like people than to things that look like spiders or cars with arms or whatever else robots might look like.
There’s also the possibility that some jobs require workers that have some human-like parts, but which human-like part depends on the job: In that case, something human-like might be able to do all of those jobs, providing flexibility.
Perhaps the oldest profession known to man? /s
But in reality, I think all industrial activities have been built around our human form, even when it’s not really the most efficient. Perhaps in farming, where some crops are sensitive to machines and humans can select the proper item? Grapes come to mind - humans can harvest wine grapes at the proper time based on characteristics of the crop, and without damaging the vines - I know there have been attempts to automate this, but not sure how successful that has been. Just one example, maybe there are others.
Limiting it to physical work, I can imagine that plumbing, carpentry and electrician work might be someday done by AI robots. Who wants to contort themselves under a dripping sink with a wrench, after all? But I expect that this will be one of the harder problems to solve: it needs adaptive thinking and improvisation to solve problems. I don’t think we’re close to that yet.
I’m very puzzled by the obsession with humanoid robots. Japan seems to have a peculiar fascination with those, for some reason: Asimo, for example.
There seems to be a disconnect between the idea of body and the brain. The interesting part is the brain, surely. The body is just cosmetic, for most purposes?
That’s a brain problem, not a form problem. Presuming that we get the software figured out, would it be better for the robot to be human-shaped? Or would it be better for the robot to roll along next to the vines on wheels while eight rubberized arms pick the grapes?
Being done with transmitted light and optical sensors coupled with machine learning to judge which grapes should be picked. Right now with “humans in the loop” virtually providing training data.
Details with a picture of the robotic system below. Definitely not humanoid.
I suspect it’s just a phase - the proof-of-concept phase. Once the robots have proven themselves capable of doing human things humanly, then we may see non-humanoid form robots doing some of the things being suggested here that today ONLY humans can do.
My assumption is a lot of robots won’t be bipedal with 2 legs and 2 arms. they will probably have 3-4 legs with wheels for feet and however many arm appendages as needed. When they are on flat surfaces they can drive but when they need to walk they can lock their wheels to turn them into feet so they can climb up/down stairs or climb over obstacles. Having 3-4 legs would provide more stability than having 2 legs also and make picking things up, putting things down, maintaining balance, etc easier.
Long term thats probably the main form. I think we are just in the early phases where we unconsciously assume the human form is superior and we want to build robots for that. But after we do that, we will realize that robots that can seamlessly transition between driving and walking and that have 3-4 legs provides the best of all worlds (high speed, mobility, low energy usage, ability to climb stairs and navigate obstacles, ability to balance, etc).
There will also be things like exoskeletons to help humans maintain stamina while working in factories, along with augmented reality to help humans work with advanced technology with less training.
I was going to say building custom houses, but that only applies to the current framing paradigm. 3D printed houses exist, and although they are VERY primitive right now, the possibilities are endless.
3D printed buildings do exist, but I doubt that’ll ever become the standard way to build a building. Just setting up the printer requires close to as much work as constructing a conventional building, and then the print process is very slow (wet concrete doesn’t support a very thick layer, and then you have to let each layer cure).
But I also don’t see why a robot to do conventional construction would need to be human-shaped. It probably would need legs, to traverse unfinished surfaces, and arms, to handle frames, beams, and other components, but there’s no reason why it should necessarily have exactly two of each.
I seem to remember reading that the original purpose for Asimo and similar humanoid robots was as a home health aide, so that, for example, the robot could pick up a disabled person off the bed and transfer them to a wheelchair or a commode.
And I saw YouTube videos of robotic arms folding laundry. That’s another thing where I imagine that a humanoid form helps (but I’m sure someone will be able to explain how a non-humanoid robot can do it better).
1: Replace the legs with wheels. That makes everything easier.
2: Add a built-in table to the front of the robot.
3: Give the robot larger hands (possibly including the table).
4: Give the robot more hands and longer arms, so it can fold sheets and blankets, too.
So arms and hands but wheels instead of legs means it’s no longer a humanoid robot? Good to know.
Um, yes? What do you think “human-shaped” means?
See, I thought things like this were still humanoid.
Vaguely, but that’s not even close to the optimal shape for a laundry-folding bot. You know those folding plastic panels you can use to make it easier to fold clothes neatly? That’s what a folding-bot would have for hands. At least, for some of its hands.
Perhaps giving the robot hands and arms would let it do other things in addition to folding laundry?
Childcare robots should probably be very much humanoid for obvious reasons.
Not sure? We are not replacing the parents with childcare providers; we are supplementing them. No reason they couldn’t be in the form of a furry or fuzzy fantasy animal that can talk?
A great deal of household work is likely to be better done using a humanoid form for the simple reason homes are built for the use of humanoid beings. We aren’t going to modify them to be built for the use of robots instead of ourselves. A robot that can climb stairs, step over junk on the floor and fit into human-shaped spaces will do better at home than one which can’t.
Robots designed to occasionally be teleoperated would also often benefit from a partial or completely human shape to better fit their operators.
Of course we are replacing them, parents have less time for their children (or anything other than work) all the time. From the viewpoint of society children are just the unfortunately necessary prerequisite for future workers, police and soldiers.