5 ms·

> But humanoid robots are much different, and much more complicated. With rockets, well, we already had rockets. And with electric cars, we already had cars, ba

by elurg 5y ago

> But humanoid robots are much different, and much more complicated. With rockets, well, we already had rockets. And with electric cars, we already had cars, batteries, sensors, and the DARPA competitions to build on. I don't think this line of argumentation holds up very well: we already have decades of research on humanoid robots too. They're just not good enough to be economically viable but they same could be said of electric cars and reusable rockets. Why couldn't Tesla just hire enough people to replicate the published research and then improve it? The total amount of money invested in robotics worldwide is probably not even that high. The capital required to duplicate Boston Dynamics doing parkour is probably much less that what Tesla is already investing in custom AI hardware.

alpaca128 5y ago
But the question is, why build a humanoid robot? It's much harder to balance (especially if it should be extendable with modules), probably won't be able to handle harsher terrain than four-legged models in the foreseeable future, and I don't see how they can do tasks significantly better than a simpler model with a few robot arms. The stature isn't the main reason for the versatility of humans. I guess humanoid robots would be useful in situations where familiarity is important, but Tesla probably wants more than just a premium customer greeter with integrated voice assistant.
- scoopertrooper 5y ago
  Our indoor environments are built for humans. For instance, robots with wheels would struggle with stairs. Short robots would have trouble reaching up and grabbing things from shelves. Wide robots would struggle to get down narrow hallways.
  - alpaca128 5y ago
    The four-legged Boston Dynamics robot can walk up/down stairs, is pretty slim and has a robot arm able to reach a bit higher. And there's still a lot of room for improvement. Even humans can't always reach the highest shelves without a ladder. I'd like to see a humanoid robot successfully climb one up and down while holding an object; that alone would be an insane achievement. A good humanoid robot costs a ton, to a point where I doubt it'll see much use in normal households. I'd love to be wrong in that aspect, but I haven't yet seen an affordable robot beyond hobby project standards. Imho a humanoid body would be a nice bonus once everything else is working reliably and there's enough demand, not what I'd start out with.
    - scoopertrooper 5y ago
      Fair points. However, I think the biggest problem with a short robot would be people tripping over it. Even if it had a big 'ol arm sticking out like a robotic giraffe, it'd be hard to gauge the dimensions of the base, especially given it may not be clear what direction it is facing if the arm can reticulate about as it pleases. So, a robot in a human environment must be about human height and width, but also its width should be roughly uniform from top to bottom like a human. I suppose you could build a quadruped of those dimensions, but it's hard to imagine what it'd look like. I am not aware of any creatures like that occurring in nature, likely with good reason.
      - MauranKilom 5y ago
        > However, I think the biggest problem with a short robot would be people tripping over it. Somehow that has not turned out to be a major problem with the closest thing we already have in widespread "use": Dogs. I think we can agree that humans are pretty good at not stepping on a retriever-sized thing by accident. Not to mention that these robots will probably not be very quiet (servos can be pretty noisy). So if that's the biggest problem, we'd be in pretty good shape.
  - throw0101a 5y ago
    > For instance, robots with wheels would struggle with stairs. Just ask the Daleks.
- regularfry 5y ago
  The reason you want humanoid robots is because environments that have humans in them are generally built around human ergonomics.
- rlt 5y ago
  > why build a humanoid robot? For one, humanoid robots are a bit less creepy than the alternatives. Have you seen the BD robots? That’s not as important for industrial use, though.
scoopertrooper 5y ago
A self-driving car is just an intrinsically easier problem to solve. In fact, the first car that automatically centred a car within a lane using cameras to track lane markings hit the road in 1977! They had cars doing cross-country trips on highways with traffic in the 1990s! Obviously there are complexities to it, which is why don't have any level 4 vehicles is just so much lower. Humanoid robots are just several magnitudes harder because you first have to master the basic mechanics of human-like movement before you can move onto higher order things like avoiding running into a person. With a bipedal robot, a simple operation like walking up a hill is a major achievement. In the car domain, you essentially get that stuff for free.
- elurg 5y ago
  Humanoid robots might be hard but they've been successfully demonstrated. It might also be argued that machine learning can make complex motions easier than if everything is hand-coded. The humanoid form factor might be an entirely unnecessary distraction, a robot with more legs or non-human articulations would be just as useful.
  - 542458 5y ago
    Boston dynamics has found machine learning to be minimally useful here, and has had better results hand-coding most things. That’s not to say that is necessarily correct for all scenarios, but BD has lots of extremely smart and motivated people and I would take their results seriously - so at very least it doesn’t seem to be as easy as throwing a big neural net at it, and any machine learning approach will require hard and careful work. Re: successfully demonstrated, again Boston Dynamics might be the best example. Their bipeds are nowhere near the functionality levels Elon is describing - fine motor skills, vision, even walking (it’s my understanding that Boston Dynamics has to individually tune their bipeds for every course) are years and years away from hitting Elon’s goals at current rate of progress. Furthermore, the BD bipeds have a battery life of about ten minutes. You need a revolution in either energy storage density or kinematic efficiency or more likely both to get a useful assistant bot out of it.
    - 5y ago

by elurg 5y ago

alpaca128 5y ago

But the question is, why build a humanoid robot? It's much harder to balance (especially if it should be extendable with modules), probably won't be able to handle harsher terrain than four-legged models in the foreseeable future, and I don't see how they can do tasks significantly better than a simpler model with a few robot arms. The stature isn't the main reason for the versatility of humans. I guess humanoid robots would be useful in situations where familiarity is important, but Tesla probably wants more than just a premium customer greeter with integrated voice assistant.

scoopertrooper 5y ago
Our indoor environments are built for humans. For instance, robots with wheels would struggle with stairs. Short robots would have trouble reaching up and grabbing things from shelves. Wide robots would struggle to get down narrow hallways.
- alpaca128 5y ago
  The four-legged Boston Dynamics robot can walk up/down stairs, is pretty slim and has a robot arm able to reach a bit higher. And there's still a lot of room for improvement. Even humans can't always reach the highest shelves without a ladder. I'd like to see a humanoid robot successfully climb one up and down while holding an object; that alone would be an insane achievement. A good humanoid robot costs a ton, to a point where I doubt it'll see much use in normal households. I'd love to be wrong in that aspect, but I haven't yet seen an affordable robot beyond hobby project standards. Imho a humanoid body would be a nice bonus once everything else is working reliably and there's enough demand, not what I'd start out with.
  - scoopertrooper 5y ago
    Fair points. However, I think the biggest problem with a short robot would be people tripping over it. Even if it had a big 'ol arm sticking out like a robotic giraffe, it'd be hard to gauge the dimensions of the base, especially given it may not be clear what direction it is facing if the arm can reticulate about as it pleases. So, a robot in a human environment must be about human height and width, but also its width should be roughly uniform from top to bottom like a human. I suppose you could build a quadruped of those dimensions, but it's hard to imagine what it'd look like. I am not aware of any creatures like that occurring in nature, likely with good reason.
    - MauranKilom 5y ago
      > However, I think the biggest problem with a short robot would be people tripping over it. Somehow that has not turned out to be a major problem with the closest thing we already have in widespread "use": Dogs. I think we can agree that humans are pretty good at not stepping on a retriever-sized thing by accident. Not to mention that these robots will probably not be very quiet (servos can be pretty noisy). So if that's the biggest problem, we'd be in pretty good shape.
- throw0101a 5y ago
  > For instance, robots with wheels would struggle with stairs. Just ask the Daleks.
regularfry 5y ago
The reason you want humanoid robots is because environments that have humans in them are generally built around human ergonomics.
rlt 5y ago
> why build a humanoid robot? For one, humanoid robots are a bit less creepy than the alternatives. Have you seen the BD robots? That’s not as important for industrial use, though.

scoopertrooper 5y ago

A self-driving car is just an intrinsically easier problem to solve. In fact, the first car that automatically centred a car within a lane using cameras to track lane markings hit the road in 1977! They had cars doing cross-country trips on highways with traffic in the 1990s! Obviously there are complexities to it, which is why don't have any level 4 vehicles is just so much lower. Humanoid robots are just several magnitudes harder because you first have to master the basic mechanics of human-like movement before you can move onto higher order things like avoiding running into a person. With a bipedal robot, a simple operation like walking up a hill is a major achievement. In the car domain, you essentially get that stuff for free.

elurg 5y ago
Humanoid robots might be hard but they've been successfully demonstrated. It might also be argued that machine learning can make complex motions easier than if everything is hand-coded. The humanoid form factor might be an entirely unnecessary distraction, a robot with more legs or non-human articulations would be just as useful.
- 542458 5y ago
  Boston dynamics has found machine learning to be minimally useful here, and has had better results hand-coding most things. That’s not to say that is necessarily correct for all scenarios, but BD has lots of extremely smart and motivated people and I would take their results seriously - so at very least it doesn’t seem to be as easy as throwing a big neural net at it, and any machine learning approach will require hard and careful work. Re: successfully demonstrated, again Boston Dynamics might be the best example. Their bipeds are nowhere near the functionality levels Elon is describing - fine motor skills, vision, even walking (it’s my understanding that Boston Dynamics has to individually tune their bipeds for every course) are years and years away from hitting Elon’s goals at current rate of progress. Furthermore, the BD bipeds have a battery life of about ten minutes. You need a revolution in either energy storage density or kinematic efficiency or more likely both to get a useful assistant bot out of it.
  - 5y ago