I'm so confused why you would use a laughably bad human when you could have a specialist robot. These things are going to get outperformed by the latter even more so than humans would.
I'm not surprised at all they're struggling to find buyers.
I'm not surprised at all they're struggling to find buyers.
Less efficiently, sure, but for the manufacturing, logistics, maintenance? The economies of scale are immense.
The reason why we weren't doing exactly that back in the 80s isn't that universal humanoid robots somehow weren't desirable. It's that for a universal humanoid hardware to be useful, you need a fairly universal AI to back it.
That "universal AI" was nowhere to be seen back in the 80s, or the 90s, or the 00s. Now, we finally have a good idea of how to build the kind of AI required for it.
Like, Digit costs a quarter mil and is rated for 10 thousand hours. It can stack boxes. For that price you can turn every box in your warehouse into an AGV and they'll last you forever.
It's impossible to do this in a general way. This could theoretically be scalable (produce the robot and have 10,000 companies all develop their own specialization routines), but the hardware (both the parts as well as neural interface) needs to be as capable as a human body, which isn't even remotely true. The physical robot will always limit what skills it can learn, on top of the difficulty of programming the skill.
I think we're hundreds of years away from making a robot that's as capable as a human. We would get there faster with synthetics or cyborgs. Create a human body without a brain, use Neuralink to operate it. Until then, specialized robots are the only thing that will scale to 10,000 skills.
Currently, dedicated robotics datasets are pathetic - in both the raw size and domain diversity - compared to what we have for generative AIs in domains like text, sound, video or images. So adding any more data helps a lot.
If you trained a robot to fully strip down a specific e-scooter model - whether for repair, remanufacturing or recycling - that training data would then help with any similar tasks. As well as a variety of seemingly unrelated tasks that also require manual dexterity, manipulation and spatial reasoning.
Those "9999 specializations" all overlap in obvious and subtle ways - and they feed little bits of skills and adaptations to each other. Which is why a lot of the robotic companies are itching to start pushing the units out there as soon as they are able to perform some useful tasks. They want that real world training data.
It's basically a robot arm with mobility at that point, and if you need more than one, just have more than one robot wheel into place. There's no particular reason to have two arms.. one, or three, or five are all sensible numbers. Heck, a chassis supporting a variable number of arms and other appendages (sensors and so on) is plausible, and the control system looks more like an ant-colony mind than a human one.
Which is a long-winded way of saying, there's no particular reason to link embodiment and cognition at the individual arm level in a factory scenario.
What remains is all the weird and awkward automation-resistant tasks where "just get a human to do it" is still easier and cheaper than redesigning everything to maybe get old school automation to handle them.
This is the kind of niche humanoid robots are currently aiming at. It's no coincidence that at least 3 companies trying to develop humanoid robots have ties to vehicle manufacturers.
Like is there any particular reason for it to be about 6' tall with exactly two 3' long, three-jointed arms rather than any of the other possible permutations for those things?
Just look up how vehicle interior assembly is performed now. Look at all the things that are still done by humans - all the different assembly stations, all the loading and unloading, all the installation operations, all the panels and wiring harnesses and plugs and bolts.
Then try to come up with a robot frame that would do all of it - every single operation that's currently done by a human - while being significantly less complex than a humanoid frame.
The design space constraints would choke the life out of you.
This just occurred to me: do standard industry robotic arms not fit that description perfectly? They're not specialized for any particular task, the only customizable parameters are the size and the end effector.
They can move around car bodies or seats, or pick up an airbrush. They can probably be installed with a five-fingered hand, or onto a giant human torso, should such tools somehow made sense for some applications. They feel like the generalist robot that meets most of the expectations for the hypothetical factory humanoids, sans being a humanoid. I mean, I get it, but aren't those existing bots just what "the vision" calls for?
The premise itself seems bogus though - there's plenty of tasks such as traditional assembly line and conveyor belt automation where a stable robot bolted to the floor, with a wired power source and custom manipulators is going to be a much better option.
For a mobile robot stability and reliability are key, and it's hard to see how a humanoid robot would be anything other than a massive downgrade for applications like Amazon's warehouse robots, hospital drug delivery robots, mall security robots, robot vacuum cleaners, etc. Wheels for the win.
OTOH there's the dream/hype of a domestic robot doing all your household chores, where humanoid form might actually be a plus, but at this point that's a pipe dream, and I seriously doubt many people really want C-3PO in the kitchen washing the dishes even if he is managing to do it without breaking anything or short-circuiting himself. It's like a 60's vision of the future, with people in flying cars or living on mars. No product-market fit.
If they’re humanoid then they can already use tools, equipment, and access methods we already use for ourselves.
What part of the vision doesn’t make sense?
When these things can make a burger without help I'll change mind, but right now they're not even close to that. Everything I've seen so far makes them look like clumsy pieces of junk. I haven't even seen one make a sandwich without a human having to prepare every step for them so they could then perform "cutting motion" or "stack ingredients" (painfully slowly and shaking like a geriatric).
These aren't. They're not even ten percent there. I don't get why you'd try to mass-produce and market them.
Tesla is going to have proper autonomous driving in their consumer vehicles before they make one useful humanoid robot.
The recent clip posted by Marc Benioff was...painful. It took a few seconds to reply to a simple greeting. Its next bit of speech in response to a query of where to get a Coke has a weird moment where it seems like it interrupts itself. Optimus offers to take Benioff to the kitchen to get a Coke. Optimus acknowledges Benioff's affirmative response, but just stands there. Then you hear Musk in the background muttering that Optimus is "paranoid" about the space. Benioff backs up a few feet. Optimus slowly turns then begins shuffling forward. Is it headed to the kitchen? Who knows!
The reaction to that should not be "OMG I cannot wait to pay you $200-$500k for one of these!" It should be "You want HOW MUCH for THIS? Are you nuts?"
- No pain
- No breaks
- No protesting/strikes
- No rises needed
- No happiness to take care of
All things business find annoying.
They could release a software update and disable your entire workforce unless you agreed to pay more money. They could slow your workforce down to prop up a competitor, etc.
Your current cloud provider can absolutely "release a software update and disable your entire workforce unless you agreed to pay more money". The reason why they don't is quite obvious (competition).
That was my entire point!
Generalist robotics are all about minimizing or at least front loading some portion of retooling cost, minimizing overhead associated with safety and compliance, and being able to capitalize what would have otherwise been human opex. Those pressures aren't going anywhere.
I have very little doubt that they’ll be making burgers within 10 years. The hardware is already capable I think - it’s mostly just a software problem now.
Their extremely clumsy fingers/hands can pretty much just open and close. I have doubts they could even fasten a screw using any kind of screwdriver, given the hardware's overall lack of precision and degrees of freedom in their digits.
They have about enough degrees of freedom to grip most things, but not enough to manipulate (such as rotate) something like a screwdriver they are already holding. Some sources claim they can be precise to within around 1 newton with their grip force - this does not include the additional error introduced by the rest of the robot the hand is attached to. This is relevant if they're holding something that is also pressed against a cutting-board. Best case for them is simply holding an object without having to exert further control, since they have some sort of clutch mechanism that decouples the motors and essentially locks their fingers in place. That gets rid of the shaking you'd otherwise see when they have to exert continuous motor effort. You do not have that luxury when you are manipulating something soft, such as when cutting a tomato, or when you're trying to turn a screwdriver in your hands.
We can deduce they are at least an order of magnitude off human precision, likely multiple. These things are not going to cut tomatoes into something you'd put onto a sandwich or burger, then sell to customers, any time soon.
What does them being humanoid have to do with this? There are other form factors that could get to 80% but might be simpler to implement.
If that ends up being a dominant or niche part of the robot market is way too early to predict.
Can that "other form factor" climb stairs? Or operate existing power tools? Or get into a generic car to get transported to a new workplace? Or get teleoperated by a human with mocap gloves?
Non-humanoid robots don't get simplicity for free. They have to trade off capabilities to get there.
> Or operate existing power tools? Or get teleoperated by a human with mocap gloves?
Requires grippers that can hold in a similar way to human hands.
> Or get into a generic car to get transported to a new workplace?
My dog can do this and traverse a flight of stairs, and she is decidedly not humanoid.
I fail to see how that would be any less complex than making a humanoid frame in the first place.
The conversation is a little reminiscent of "before the car was invented, if you asked what people wanted they would have said a faster horse". If robots became popular in day-to-day, it's not hard to imagine that we would make space for them in our lives anyway. Cars can't traverse the kinds of terrain a horse can, and they require fuel so we have roads and gas stations. If you made a robot that was actually helpful and couldn't take the stairs, you'd start installing dumbwaiters in buildings.
Robots that can't solve a diverse range of tasks in arbitrary human-made environments aren't going to "become popular". And if robots don't "become popular", no one is going to redesign every single environment to suit them better.