Even Elon Musk style robots puppeted for pennies an hour by gig workers and AI wouldn't be very cost effective.
A useful home robot needs to have about the same level of force available to complete tasks. The range of side effects is unlimited, and there is no culture of safety, and there are kids and pets in the mix. You're just asking for trouble that really can't be avoided through technical means.
Static installations? Sure. Glasgow uni robotics demonstrated washing folding, it's a good test of complex tasks.
The reason is mixture of hardware and software constraints. You need a range of sensors and equipment (end-effectors, batteries, GPUs), expensive at lower volumes, to extend the robot's physical capabilities (e.g. reach, manipulation, navigation) and enable certain software robot skills. Besides their dependence on hardware, robot skills are not entirely solved nor general enough to work in all environments, that means the company needs to do R&D and data collection, or purchase bought elsewhere. For example, Generative AI models (LLMs, VLAs, world models) are a boon for robotics thanks to knowledge reuse and eased domain adaptation but they're (for now) somewhat unreliable. It's difficult for such embodied GenAI models to be more than technically correct when performing tasks because they lack or ignore knowledge about the physical world needed to ensure risk-free actions and outcomes.
For example, asking for a robot to "pour water on that glass" can lead to dropped bottles/glasses or water pouring on a table because the model won't have a clear models of bottle/glass/water ("entities") nor expectations (nothing broken, nothing wet; only what is more or less expected with the act of pouring water conditioned on the most probable areas for representing the of object of interest.
Just have a look at 1X's videos, a well-funded humanoid robot startups, and pay attention to object interactions: how those interactions start and end.
IMHO the biggest obstacle is that AI is still having trouble with object permanence and real world interactions. I'm hopeful this will be solved in the next few years, and I'm sure teams of people are working in it with vast budgets of DARPA dollars and COSTIND Yuan as well as private industry.
Elon could do it if he could just invent a time machine first.
Because humanoids aren't a great form factor for robotics.
Robot vacuums, washing machines, dishwashers, "self" driving cars...these are all consumer robotics and are quite popular.
Let's say I wanted a robot to take out my trash. It sounds simple but there are so many incredibly difficult tasks when you break it down, each with a near-infinite number of variations in different homes:
- First, learn where in the house it is, and how to get to it. * Is it in a drawer? What kind of drawer, how to open it? * Is it a plastic garbage bag in a bin, with a foot lever? In a drawer? - How does the robot lift out a plastic bag, replace the plastic bag? We don't have the dexterity to do this yet * What happens when the plastic bag gets caught slightly on a corner, or begins to rip? - Let's say we pick up the plastic bag, now we need to move to a door that will take us out of the house * Are there stairs, pets, children, other obstacles that could get in the way? Just this bullet point here could be harder than self-driving cars, which is far from solved * How do we interact with the door to open it? Is it a round knob, is there a deadbolt to unlock, does the door swing outward or inwards?
...etc
This probably barely scratches the surface on all of the variability inside of a home, and yet a small kid can do all this without even thinking, while even a small subset of these problems probably requires billions of dollars to solve in a controlled/closed environment.
Maybe humanoid robot teleoperation + artificial intelligence will get us there, that's the pipe dream of a lot of these humanoid robot companies. But then they need to make money and out-compete some young/lower-skilled workers happy to do things for $10/hour. At which point one wonders how these companies will make money to justify the insane R&D needed for even the simplest of tasks. But hey the same story has played out in other industries where robots have outplaced low-skilled labor. The difference though is that these environments have been heavily controlled, i.e. the same few steps to assemble a widget, the same motions to clean the same type of object, etc.
