Back in December 2020, Hyundai purchased an 80% controlling interest in Boston Dynamics from SoftBank for $880 million, part of a transaction that valued the robotics company at $1.1 billion. That agreement included a put option allowing SoftBank to sell its remaining stake to Hyundai at a later date.
Oh. It's Softbank exiting humanoid robotics at Softbank's discretion. That's a lot different than " Hyundai buys Boston Dynamics". Hyundai bought them years ago. This is just the last 8%.
Seems like a mistake. AI in its current form has limited usefulness for most people. Not something I would pay for to use outside of work. But a household robot that could clean, wash and fold the laundry, do the dishes, maybe even be a chauffeur... that would be huge. I think a lot of people would pay new-car money for something like that.
> AI in its current form has limited usefulness for most people.
That's not what I'm seeing. My mom always wanted Google to just answer questions, and now ChatGPT can. She uses it enough in her daily life that she bought a subscription.
Yes, she knows it hallucinates and you have to double check everything, but so far she finds a ton of use for it even with those caveats.
Now, I agree that a personal servant robot would get a ton of business. Even at new-car prices, it's still cheaper than a human caretaker/maid/butler/etc. And the maid usually doesn't also mow the lawn on a hot day, while a robot would potentially do all kinds of different things without complaint.
> robot maid that could clean, wash and fold the laundry, do the dishes, etc. would be huge. I think a lot of people would pay new-car money for something like that.
Once you take maintenance of a machine with price-parity to a new car into consideration, it’s surely cost competitive to just hire a human to do all those things.
The price needs to fall drastically below new-car territory before it’s competitive with manual human labour.
> ”it’s surely cost competitive to just hire a human to do all those things.”
The cost of labour varies hugely in different parts of the world. The cost of hiring someone in Switzerland is on the order of 100X more expensive compared to Bangladesh, for example.
With many countries currently in an anti-immigration political mindset and with birth rates declining globally, labour costs are likely to continue to increase in the future.
But once a technology like general-purpose humanoid robotics exists, costs are only likely to decrease over time.
Humans are messy to deal with. Say you're rich enough to afford a personal chef. Unless you're an inhuman monster, their problems become your problems as well. So if your chef is out because their mother is sick and needs someone to take care of her, you pay for a nurse for your chef's mom, so that you have your chef. A robot servant is still gonna need maintenance, sure, but it's a bit easier to be callous to a robot than a person.
But for now, humanoids are still relatively incompetent. So if you're rich enough and you want things done like cleaning and cooking, it's more convenient to hire a human and give them specific instructions instead of buying a robot that can currently do only half the things they want to get done. Or they just want a nice gadget.
If you're paying cash/under the table, then maybe. But even then a twice a week household cleaning hire is going to cost upwards of $1500/mo unless you're being particularly exploitative. If you're not under the table, you're paying payroll taxes, probably paying for a payroll service, etc. so you're talking $2000+. At best you can maybe stay under $20k a year.
When you really look at the economics of it, a robot that never gets sick/doesn't require payroll/etc. makes a lot more sense.
This comment seems insane to me. Like at $50 an hour thats 30 hours a month, or 8 hoursish a week. How dirty or huge is the house? And $50 an hour is way over what most hourly degree edgucated workers earn so definitely not exploitation.
> But even then a twice a week household cleaning hire is going to cost upwards of $1500/mo unless you're being particularly exploitative.
Sorry, what? Unless you're doing a deep clean of your house twice a week or you live in a particularly HCOL area, those numbers don't add up. You shouldn't be spending more than $1k/month on household chores, and even that seems high.
Source: A client of ours runs a "personal help" service (mostly focused on household tasks like laundry, tidying, organizing, etc as opposed to deep cleaning) so I have a lot of data on this. And they're a relatively premium service compared to some of the cheap labor you can actually buy. But they also don't operate in SF or NYC, so maybe prices are drastically different there.
Who spends new-car money to clean their homes? Maybe ultra high net worth individuals? I know people with 8 figures net worth who spend a fraction of that money for cleaning their homes.
> But a household robot that could clean, wash and fold the laundry, do the dishes, maybe even be a chauffeur... that would be huge.
Except they can't. I get it, merging advanced AI with robotics has made huge leaps in the past few years, but building a truly autonomous laundry robot is an incredibly difficult problem that still feels many years away. And I've seen all the "folding robots" over the past few years, and they are still miles away from being useful in your average home (they only "fold" if pieces are handed to them one-by-one, or the more advanced ones that can pick out clothes from a pile look like they were folded by a 3 year old).
Also, consider that all-in-one washer/dryer combos have existed for a while, but they are still a teeny percentage of washer sales because they're expensive and require more maintenance. There is a surprisingly low threshold on what people are willing to pay for labor saving devices.
Getting to a cheap household helper robot requires building the expensive one. I don’t think Boston dynamics believes any reasonable consumer would buy their atlas robot, but by building it and scaling its production in industrial use cases they will learn things that make building a cheaper one much easier. And they will have built some factories to mass produce them. It’s not something that will be in everyone’s home next year, but sooner or later the robot hardware and robot intelligence will both be cheap enough to be accessible by average people (at least in the developed world)
> There is a surprisingly low threshold on what people are willing to pay for labor saving devices.
A lot of people gladly pay humans to launder their clothes, so the market is there. But the current iteration of wash/dry combo machines doesn’t solve the main issue. I don’t mind transferring my clothes from the washer to the dryer, because that takes 60 seconds. It’s folding and putting away the clothes that makes laundry a chore.
> Not something I would pay for to use outside of work
You wouldn't but apparently your employer would.
I don't disagree with you on robotics, though. For an empire like softbank, not buying an "insurance against the rise of robotics" also seems like a mistake to me.
That being said, they may expect robotics to rise through self-driving cars (hence their investment in Wayve).
Is there any sign beyond flashy demos that humanoid robots will be functionally feasible though (before we even get to economically feasible)?
I know there's tons of activity on humanoid teleoperation data collection, and motion model training, but it hasn't seemed to bear out much of anything.
Like.... AI would be great if I could put it into a magical semi-corporeal familiar but I'm just not seeing a path to those either.
Well, that assumes that if you just keep throwing more data and compute at large language models you'll end up with something akin to AGI to control those robots. Which is far from guaranteed.
LLMs aren't the specific architecture you'd use, but it very much looks like a tractable engineering problem to go from a university research lab project that can manage to fold clothes as a demo, to a sellable consumer product. The timeline is gonna be off, so no one knows if it's gonna take 3 years or 30, but it's not going to take an unknown breakthrough in materials science and physics the same way that nuclear fusion looks like it will require.
Now how about a household robot that does all those things and is controlled by an Elon Musk company or by some other completely benevolent techno oligarch?
$325M for 9.65% implies a valuation of around $3.4 billion, so it's more like 18x.
Regardless, Boston Dynamics has been burning cash for 35 years and all they have to show for it are some fancy demos and trial deployments. Eventually you have to wonder what their future prospects are.
I was trying to think about the why with Cursor, and the only thing that makes sense to me is they wanted experts in making harnesses so that they can pivot that expertise towards building harnesses intended for autonomous agents to use instead of humans. There's no world where a 60 billion IDE makes sense.
I think it's more that Space X's valuation is ridiculous, and they need acquisitions that are ridiculous to pretend that the emperor is still wearing clothes.
When cursor is selling for $60B, then grok has to really be worth several times that, right?
> This was just ~10% of boston dynamics at that price. HN pro-tip: before commenting, read the articles not just the headlines.
I'm not sure if you're following your own advice...?
The ~10% just sold was bought for $325 million.
The total price they paid was $1.205 billion ($880 million in 2021, $325 million now).
The $1.1 billion figure in the HN headline is kind of just wrong and presumably based on what they considered to be Boston Dynamics' total valuation in 2021, but represents neither what they paid for the ~10% nor the actual total they paid over both transactions.
I don't understand why they would implement humanoid robots instead of purpose-built robots. The human form is not the most optimal way to do most tasks, especially as it relates to manufacturing. Robots don't need to look like humans, they need to be useful. Seems like putting in an awful lot of extra unnecessary work to end up with a worse result.
I'm not sure how many times this has to be restated.
It's car manufacturing. Everything that could be done by a purpose specific robot arm bolted down to the factory floor is already done by a purpose specific robot arm bolted down to the factory floor.
What remains unautomated, then?
The long tail of tasks that are too minor, too finicky, too open-ended or too reliant on manual dexterity to be offloaded onto traditional robots.
This is where this new generation of robotics comes in. This is the kind of task they're designed to do: "a task that's still done by a human in a high automation environment". Universal robots are angling for the tasks that are impossible or uneconomical to automate with traditional industrial robots.
> Everything that could be done by a purpose specific robot arm bolted down to the factory floor is already done by a purpose specific robot arm bolted down to the factory floor.
Hah! Hardly. I say this as someone whose first "real job" was in applying robotics research to automotive assembly - there are still a ton of assembly tasks that could be performed by a fixed-base robot arm, or a robot arm on a linear rail/fixed gantry. Wheeled mobile manipulators are only needed in a few cases, and humanoid form-factor is only "necessary" in very few cases (and I don't think the current crop of humanoids is particularly suited to these tasks).
In my opinion/experience, the impediments are that (1) the system integrators that are usually responsible for assembly-line robotics are too stupid to figure out how to apply robots to the problem, (2) the automakers themselves are often too short-sighted/stupid/unwilling to invest in increased automation (and particularly in building the in-house competency that they really need), (3) the hostile/exploitative relationship between (most) automakers and their main suppliers means that low-hanging improvements to parts/assemblies are a non-starter, and (4) the automaker C-suite (and investors) are too drawn to silver-bullet solutions (e.g. humanoids) than practical automation improvements.
"Could be in principle" and "could be in practice, under technical and economical considerations in play" are two very, very different beasts.
Everyone in the industry learned that the hard way.
At a certain point, the tasks that remain stop being "dexterity" problems and start being "AI" problems. That is: a robot could do the task - if you either spent big $$$ on redesigning the entire task around the robot's intellectual limitations (uneconomical), or if you had an incredibly advanced AI capable of problem solving driving that robot (impossible with 00s AI).
The "universal robot" bet is the "incredibly advanced AI capable of problem solving" bet. That in 2020s, AI is finally capable. The body only has to be "good enough to make most tasks possible".
> "Could be in principle" and "could be in practice, under technical and economical considerations in play" are two very, very different beasts.
> Everyone in the industry learned that the hard way.
The auto industry is notorious for making incredibly myopic choices to save money/make money in the near term versus long-term investments. The relationship between automakers and their suppliers/vendors is basically a century-plus of the automakers trying to (1) outsource anything they can for a quick buck, and (2) grind the supplier/vendor margins down to nothing. (This is part of why the newer Chinese automakers with much greater vertical integration are such a threat to the traditional automakers; vertical integration has a high up-front investment but the payoff in flexibility and speed is significant).
Vertical integration is superior if you can pull it off. Big fucking "if". There's a reason why automakers don't usually do it.
The name of the reason is: corporate rot. They don't have the organizational backbone that wouldn't let their "in-house manufacturing" rot away into inefficiency and waste.
Not that it has much to do with why automation fails to penetrate certain tasks. The reason why "long tail" tasks are often beyond automation is: piss poor ROI, calculated correctly.
You go out of your way to automate a certain process with traditional robotics, and it'll probably pay off in 15 years. The chassis this applies to is going to be in manufacturing for 10 years. At least half the systems work you've done there would have to be redone for the next chassis. Fun.
The bean counters counted their beans, and found out that using traditional robotics there is a losing game. Thus the search for better options. And the humans performing the tasks in the meanwhile.
> Not that it has much to do with why automation fails to penetrate certain tasks. The reason why "long tail" tasks are often beyond automation is: piss poor ROI, calculated correctly.
I actually don't think any of the big automakers have ever really, in-depth considered the ROI of "traditional" assembly automation (i.e. anything SoTA pre-2020), with experts in all parts of the process in same room. It's easy to assume that these companies must make careful measured decisions based on evidence, but in practice big decisions are made by small groups within the C-suite, often pretty divorced from the reality on the ground.
For example, many of the big asian automakers seem to have completely ignored the well-understood effects of their demographic crises (i.e. significantly aging population) on the future of their workforce (i.e. they are having trouble retaining and hiring new workers as the older generation retires) and this totally changes the economics of automation! Now they are all having to play catch-up, having realized that they must automate, at whatever the cost, because the issue is not "robots must be cheaper than human labor" it is "we might not be able to afford human labor at all".
I have visited factories for work and my experience is the same. There is so much stuff that could easily be automated but is not because it is too expansive for too little value to make a custom one off machine. The big high volume things will be automated but these machines will have 90% success rate and lot's of stuff that needs to be done by hand. You can search for factory tours on youtube to get an idea. Here are two videos, an Amazon warehouse and a Tesla factory. the big heavy stuff is automated but lot's of work is still done by hand.
https://www.youtube.com/shorts/-R6cBkza17khttps://www.youtube.com/watch?v=45slYC99uUg
This is over the last decade at one of the largest automakers in the world. Naturally there is significant variation between individual lines and plants; some are newer and more automated, some are older and much less automated. Are some cars being built on more automated lines? Yes. But a great many, probably the vast majority, are being built with fairly low assembly* automation.
* There is a significant split in automation between "body weld" stages and "assembly" stages. Body weld is very heavily automated basically everywhere (although there are some surprising exceptions in places), while assembly is much less automated.
“One of the largest automakers in the world” makes me think of very low-tech companies like Ford or whatever. I can’t imagine this would bring much actual experience with this new generation of robotics.
Ford doesn't even make the top 5 - and however "low-tech" you think these companies are is the point, the overwhelming majority of new cars are being built by those "low-tech" automakers. The problem is not the limits of current technology (or even of the state of the art 10 years ago), it is the lack of vision and will within these companies to invest in using it.
> I can’t imagine this would bring much actual experience with this new generation of robotics.
Luckily for you, my job has always been within the robotics research side of the company, so I am very much aware of the strengths and weaknesses of the current technology.
This gives me vibes of "... but they're dinosaurs" that pervades Tesla-type discussions. I've watched several extensive discussions here with some Tesla fans breathlessly announce "really cool", "futuristic" "new thinking" functionality or features for their cars that "the dinosaurs just aren't even considering"... except to their initial skepticism, disbelief, actually, the dinosaurs often do have those things (examples including "adaptive blind spot", where triggering of those alerts is highly dependent on speed differentials, so if the vehicle is more rapidly approaching you, the alert fires earlier, dynamic traffic sign recognition - my car doesn't just recognize school zones, but can see when they are active if they have flashing lights, and, based on equipment installation, will actually count down to when my intersection light will turn green, or road mapping radar, where it actually scans the road surface ahead, avoiding potholes when safe, and adjusting ride height dynamically, not just 'press a button or choose a mode to raise or lower').
Indeed, Ford. Leaving aside the "old school" six axis robots that have been around for decades, Ford absolutely uses UR10s collaborating with humans to sand the entire car body in about 30 seconds, and to fit shock absorbers. They're also used at the engine plants. They also use the Symbio platform for transmission assembly, and fully autonomous forklift robots throughout their Tennessee plant.
automotive workers unions started around 1918 and became major political players in the 30s -- a fact that i'm sure is wholly unrelated to why there are so many un-automated tasks in that industry.
It’s not really a secret that most new auto factories serving the US tend to open in places where those unions are not active, like the South or Mexico
Pretty different tasks, environments, outcomes, metrics, goals, and other things in a warehouse vs. a factory… really have no clue how this is supposed to be relevant. Why not mention farms or libraries?
But if these tasks are too minor, too finicky, too open-ended or too reliant on manual dexterity for a purpose-built robot, how can a general purpose robot perform them better? If anything, they should be doing worse.
The only thing I can think of are tasks that are so rarely done, it's not economical to build a robot for. But I then I also don't see how another robot solves this problem.
A) the idea is that these robots do have dexterity capabilities a lot closer to human hands
B) there’s a long tail of individual tasks it’s uneconomical to build purpose-built robots for each individual task. But it’s economical to have 1 robot that can do all of them.
Do they? A human can both chuck kilograms of stuff across a room or kick in a door, but then pick up a single hair off the ground, or feel and manipulate (things even lighter than) a literal feather.
Robots can certainly do things more repeatably, if not more precisely.
The human arms and hands are very versatile, and imitating them is a good choice for a universal robot, though 3 or 4 arms are definitely better than 2, and the hardest to imitate are the sensors, not the actuators.
But the rest of the human body is not useful in a factory environment, so the arms could be mounted on a mobile base that does not have any resemblance to a human.
And C) they don't always have to be at parity with human hands to be good enough because humans are flat out expensive. Humans need health accommodations, have sick days, vacation, and make mistakes too. The bar is much lower and the incentives are much higher than many people probably think.
These robots operate on completely different principles.
One can lift insane weights, has insane torque, and absurd precision, and can do the same movement millions of times with virtually no deviation. You program these with an exact movement plan, just like you would programm a CnC with a tool path. They are basically cnc machines.
The other one is a inacurate, unstable, dynamic system controlled by neural networks and heuristics. It has massive deviation over each run, but that means that the programming must be able to account for it. Which makes it suitable to operate on problems that are messy, unrepeatable and human-shaped.
Pretty much. It's a total paradigm shift from how industrial robots normally work. A pre-planned motion executed carefully and precisely vs open-ended "do this thing" powered by a very large bag of opaque neural network heuristics.
A robot that has to be carefully adapted and set up for the task vs a robot that you can point at a task and have it figure out how to do it. A robot that doesn't deviate vs a robot that absorbs all kinds of deviations.
It's a bet that The Bitter Lesson will win over Moravec's Paradox, in the end.
It's also worth noting that when e.g. inputs to a stage might have unpredictable defects or alignment, a robot arm utilizing neural networks for planning and analysis might still be the best way to handle that - without the extra degree of freedom of movement-relative-to-floor, planning can be done more rapidly, and movement can be executed more aggressively and quickly.
If I were Hyundai, I'd be looking at this as buying a significant amount of vision, dynamics, and integration systems expertise, not necessarily the dream of self-motive walking systems.
Well, humans obviously do those jobs, so a clearly a general purpose robot (in this case, a biorobot) has been found to do the job better. Don't overthink it.
Because it is general purpose. We did not have the ability to create a single robot form which could do all of these minor, finicky, and opened ended tasks. Now that seems within reach. The nice property of humanoid robots is that the world is already made for human form, and so if you're trying to replace people naturally this is what you'd want.
>> how can a general purpose robot perform them better
Better than what? It seems that as long as they perform the tasks "better" (cheaper / faster / lower-error) than the humans that are currently performing them, that is an improvement for the factory owner.
It's not a "general purpose" robot, it is a "human replacement" robot, with similar skills and shortcomings to a human. Humans are not general purpose.
All you need to do is look at a recent video of car manufacturing process, and watch what the humans are doing.
>It's car manufacturing. Everything that could be done by a purpose specific robot arm bolted down to the factory floor is already done by a purpose specific robot arm bolted down to the factory floor.
>What remains unautomated, then?
Stuff that can be done by purpose specific robot arms on wheeled platforms, which is very difficult, but will be much more feasible than a humanoid robot doing anything.
What you need then is a better arm (or even just hand), not a human.
Or a new take on car design with automated production in mind regarding all the wiring and what not (easier said than done, I'm sure many have tried and failed, but eventually someone will succeed).
So we want to replace a lot of cheap human labour with expensive robots?
Even if I think this has legs, where do the cheap humans go to work after? Where/what are the remaining jobs for all of this displaced labour in both white collar and blue collar worlds? It basically screams UBI. And in a UBI world, the economy looks pretty different and humanoid esque robots start to look either very altruistic or very dystopian depending on how hard the oligarchs don't want things to change.
Like, what's the end game for humans in this path we're embarking?
Naw, the real answer is that factories have been built around human labor - they weren't built to be forward-compatible with purpose-built robots, so during the transitional period where we build these purpose-built robots, you need humanoid robots to fill in the parts where the factories were geared for human labor.
'Everyone over 30 knows this' is a prior assumption (it is not necessarily correct; and nothing is said about shame).
The comic strip is saying if above is true, then people still have to learn at some point so on average it would be around 10k people per day.
I think the math is this:
For people born in a given year: 4000000/365/30 = 365 people per day
but you have 30 sets of those people (those born this year, those born last year, those born two years ago, etc.) So 365 * 30 = 10950. 10k is easier to say for viral purposes.
I took a tour of the BMW Group Plant Spartanburg body shop. It's heavily automated with industrial robots inside safety cages. But they still have human workers pick up parts from rolling carts and place them into templates for the robot arms. BMW has been running a trial with Figure humanoid robots to automate that remaining piece. Apparently those robots haven't worked very well, but presumably Hyundai thinks they can do it better?
Some colleague of mine visited the Mini plant earlier this week. Apparently they had a Boston Dynamics dog patrolling simply to spot stuff that had been left where it wasn't supposed to be left
We have a world built for humans, designed for humans to walk around and get things done. How, exactly, would it not be useful to have a robot that looks, walks, runs, jumps, lifts, carriers, pushes, pulls, twists, bends, steers, and labors like a human? It would obviously be incredibly useful.
Humanoidal robots make sense when they need to operate in spaces designed for humans bodies. Cars are designed and built to be used and serviced by humans, especially their interiors so you need humanoid robot to automate building them. Car exteriors are not built for humans to interact with so they are already being built by specialized robots.
Custom built robots are expensive (basically an R&D project in themselves) and inflexible, if you want to update the process you have to redesign your automated system. The dream of humanoid robots is they can adapt to new manufacturing processes like human workers.
Training data of task completion. See, e.g., robots doing backflips. Presumably there’s an optimal robot for gymnastics but if you start with humanoid form you can train based on many videos of human movement. The alternative - world model sim with physics and loss functions- perhaps ends up being too unconstrained when you add in optimization of the robot form…
AI watching videos of humans seems an incredibly inefficient way to solve acrobatic balance. It is just physics and engineering. The hard part is in the materials and assemblages for enabling complex and subtle movements and the fine motor control, not knowing what the motions are.
I understand the sentiment, but you assume they are planning to build humanoid robots to walk around the human-oriented space in the factory.
Perhaps they want to put some of the sensing and control features in, so a humanoid-like dexterity or adaptability for the business end of a floor-mounted robot arm?
Because a humanoid robot can replace (theoretically) any human worker, whereas a purpose-built robot can only replace one kind of worker. Or at least that's what Asimov said in Caves of Steel.
I'm spitballing here as I don't actually have a concrete answer for you. But from my understanding automobile manufacturing is one of if not THE most advanced 'purpose-built' robotics sectors. While I agree with you that having a purpose-built thing usually wins out for assembly line manufacturing, I wonder if this isn't an attempt to branch out away from single-purpose robotics into more general or multi-faceted manufacturing.
The gist of it is that all tools on the spacecraft (eg: space-drill, space-coffee-maker, space-airlock) are all designed to fit a gloved human astronaut hand. Waaaaay more complicated to make a robo-hand than a robo-suction cup or robo-claw, but then you are matching the environment, and guarantee tool compatibility against all extant tasks!
We already have specialized robots on earth... paper slicer, lawn mower, bazooka, whatever. They're all machines that are specialized for the task at hand, we're not making a humanoid robot that gets down on all fours and individually plucks blades of glass.
The car factories already have specialized robots... they're not mimicking a human hand holding a can of spray paint, shaking it up, and painting the car that way... it's a 6-axis arm, or a whole "grab the car and flip it while spraying paint" system.
It's not about inventing purpose-specific robots, it's about handling that long-tail of "stuff with tools that a human is designed to be able to use." Go over there, push that button. Go move that box from table1 to table2. Etc.
For well defined tasks in the factory domain, make a "real robot". For ad-hoc tasks in the interim... strap an LLM to a camera, battery, robo-legs and arms, cross your fingers, and hope for the best?
I think the rationale is that they are already using typical car factory automation, but they see a huge potential market for general purpose robotics in the coming decades, they don't need the humanoids, they are simply dogfooding a future product.
I think this is smart and not very risky. Tesla is playing a similar game with Optimus, for now Hyundai/Boston Dynamics is at least 5 years ahead.
That's like saying there's no need for a generic CPU. The only way forward is a ASIC. Once a generic CPU does everything well enough, it's extremely versatile.
They have god knows how manny bajillion dollars tied up in machinery designed for human use, a human can step in when the robot breaks, and they can buy more robots if the humans get uppity. Those seem like a bunch of good reasons to me.
The humanoid form factor is certainly may not be ideal but I guess they think the flexibility is worth it
You don’t understand because you’re not an expert? First off they have numerous non-humanoid robots if you follow them at all. Second, Clearly they’ve got strong reasoning, they’ve just been bought. Third, out of thousands of attempts at humanoids, their robots are seemingly the best we’ve yet seen in this class. They must be doing it right when so few others got any traction.
This same nonsensical question again. Because the world is built for humans, because these are general purpose to replace a human laborer. It can immediately go from picking up parts in a factory to mowing the lawn in the same day.
I don't think this is solely tied to car manufacturing automation. Even though Hyundai Motor Group is acquiring them, I would imagine they'd be well-positioned to commercialize general-purpose robotics and not just for car manufacturing, if Tesla is anything to go by.
I do think this might be tied to South Korea's demographics, by 2040 the working-age population is projected to decline 25% from 2020 and keep declining almost linearly until leveling out around 17M around 2065, a 50% drop total in < 50 years. I would think HMG / Hyundai sees a huge business opportunity or this might be a national-level political priority but I don't know the specifics.
Hyundai bumped their ownership up to 100%, and took the opportunity to reset expectations about when Atlas would be working in Hyundai factories.
While Atlas is the best humanoid robot so far, it still isn't useful in a car factory that's fully equipped with the latest factory robots that are strong enough to juggle car engines and are bolted to the floor.
Everyone knows the killer app for humanoid robots is building the Mars colony amirte?
There are plenty of tasks a relatively weak humanoid robot is well suited for. Basically any task that is 'human shaped' and not worth an automated line.
There's got to be something wrong at the core of BD. They've been pawned off a bunch of times, and they still don't have products out the factory line like they should. I think the tech community has been impressed by their videos, but the fact that their most sold thing is a toy dog at a luxury car price point says a lot about the company.
My personal take is that one of the reasons is their posture against ML. They've been very "GOFCT" and have only recently started to incorporate ML concepts.
>> There's got to be something wrong at the core of BD. They've been pawned off a bunch of times,
Well...there is the uncanny similarity to the T-800 and and uneasy realization that the owner of BD could become Cyberdyne Systems IRL. Perhaps some companies like that notoriety but not sure if many want that.
What is Gofct and does robotics industry generally just have had a slower adoption of ML because of the realtime domain requirements, I'm just curious and wondering aloud here.
Is it really that surprising that no-one has invested the time and effort into figuring out the personal information of some tech employee-turned-founder? I bet no-one outside of tech even knows his name.
Yeah. Google was too impatient and forced BD to productize prematurely (Spot, Handle), then dumped them when it didn't work out immediately. AI just wasn't ready yet. Imagine if Google had let BD focus on research until DeepMind was ready with the AI side of things. I think with the right joint research program they could have already been deploying humanoids today.
ai generated imagery can’t be copyrighted while all other photography can and generally needs to be treated as it is. Therefore you likely have to pay a royalty to Getty or other asset outlet. Of use AI.
There's a saying in Zen which I live by. "How you do something is how you do everything".
Start to be sloppy somewhere, you'll be sloppy everywhere. As we "learn and enable" to do things faster with less effort, the quality of the thing we (as in humans collectively) do decline.
AI, when used as the sole blunt instrument, accelerates this dramatically.
Little appreciated fact is news orgs have full time employees just dealing with licensing all day long, and they pay out millions of dollars when someone fucks up.
So true. I hear they replace more engines than any other brand. I'm surprised they sell so well; a used Toyota would be a far better choice than a new Hyundai
Although confusingly, the battery/motor on a Hyundai EV is covered under a different 10-year/100K warranty which does transfer to used car buyers. Important because of their unfixed ICCU problems.
Their much hyped 10 year 100K mile power train warranty is only for the original purchaser. Once sold the warranty reverts to a more standard 5/60K term.
Of the important life-lessons to have before one turns 18;
"don't ever be the third or fourth owner of a Hyundai" is right up there with not eating the yellow snow.
There's no shame in being broke, of course-it is merely a catastrophe. The fourth owner of a '11 Sonata is gonna have a different outlook than the fourth owner of a '73 Mercedes 600 Pullman.
So this appears to mean that Hyundai is effectively taking BD's humanoids "off the market" (B2C/B2B markets). And Softbank wants to take a different humanoids stake in OpenAI's plans.
Just a note for the thread Hyundai Motor Group makes cars and a lot of other heavy industry things - trains, defense, plant. Rolling robotics fully into their motor group makes complete sense.
rnteresting part is defense although the economics and power/range limits its practicality.
the other timing here is the increasingly belligerent union who are demanding pretty outrageous compensation for what a typical American worker would make. I think the goal is to immediately automate the workforce and move the plants out of korea speaking to insiders.
This is a bit disappointing, isn't it? Boston Dynamics had the coolest robots and everybody was marveling how they would take over the world eventually. Turns out the market isn't gigantic and the use cases are limited, at least for now.
However, let's hope they will keep on doing cool stuff under their new owner.
Outside of some military applications and maybe search and rescue, a lot of people kind of freaked out about Boston Dynamics. They have cool robots, sure, but at what cost if they are implemented by a bad actor? No thanks.
I don't think that follows. Hyundai could well sell these after they've dogfooded them for a while.
Car factories seem to be a pretty good initial market, given that Tesla is doing Optimus and Figure has humanoids in a BMW factory. But the whole point is that these are general purpose robots, and there are lots of other factories. By the time that market is saturated they'll be capable of more.
SoftBank has now exercised that option.
That's not what I'm seeing. My mom always wanted Google to just answer questions, and now ChatGPT can. She uses it enough in her daily life that she bought a subscription.
Yes, she knows it hallucinates and you have to double check everything, but so far she finds a ton of use for it even with those caveats.
Now, I agree that a personal servant robot would get a ton of business. Even at new-car prices, it's still cheaper than a human caretaker/maid/butler/etc. And the maid usually doesn't also mow the lawn on a hot day, while a robot would potentially do all kinds of different things without complaint.
Once you take maintenance of a machine with price-parity to a new car into consideration, it’s surely cost competitive to just hire a human to do all those things.
The price needs to fall drastically below new-car territory before it’s competitive with manual human labour.
The cost of labour varies hugely in different parts of the world. The cost of hiring someone in Switzerland is on the order of 100X more expensive compared to Bangladesh, for example.
With many countries currently in an anti-immigration political mindset and with birth rates declining globally, labour costs are likely to continue to increase in the future.
But once a technology like general-purpose humanoid robotics exists, costs are only likely to decrease over time.
When you really look at the economics of it, a robot that never gets sick/doesn't require payroll/etc. makes a lot more sense.
Sorry, what? Unless you're doing a deep clean of your house twice a week or you live in a particularly HCOL area, those numbers don't add up. You shouldn't be spending more than $1k/month on household chores, and even that seems high.
Source: A client of ours runs a "personal help" service (mostly focused on household tasks like laundry, tidying, organizing, etc as opposed to deep cleaning) so I have a lot of data on this. And they're a relatively premium service compared to some of the cheap labor you can actually buy. But they also don't operate in SF or NYC, so maybe prices are drastically different there.
Except they can't. I get it, merging advanced AI with robotics has made huge leaps in the past few years, but building a truly autonomous laundry robot is an incredibly difficult problem that still feels many years away. And I've seen all the "folding robots" over the past few years, and they are still miles away from being useful in your average home (they only "fold" if pieces are handed to them one-by-one, or the more advanced ones that can pick out clothes from a pile look like they were folded by a 3 year old).
Also, consider that all-in-one washer/dryer combos have existed for a while, but they are still a teeny percentage of washer sales because they're expensive and require more maintenance. There is a surprisingly low threshold on what people are willing to pay for labor saving devices.
A lot of people gladly pay humans to launder their clothes, so the market is there. But the current iteration of wash/dry combo machines doesn’t solve the main issue. I don’t mind transferring my clothes from the washer to the dryer, because that takes 60 seconds. It’s folding and putting away the clothes that makes laundry a chore.
You wouldn't but apparently your employer would.
I don't disagree with you on robotics, though. For an empire like softbank, not buying an "insurance against the rise of robotics" also seems like a mistake to me.
That being said, they may expect robotics to rise through self-driving cars (hence their investment in Wayve).
I know there's tons of activity on humanoid teleoperation data collection, and motion model training, but it hasn't seemed to bear out much of anything.
Like.... AI would be great if I could put it into a magical semi-corporeal familiar but I'm just not seeing a path to those either.
Cars will be able to safely drive themselves autonomously well before there is a humanoid robot capable of safely driving non-autonomous cars.
So why do you suddenly think you need it for controlling a body when animals do it with far less?
Regardless, Boston Dynamics has been burning cash for 35 years and all they have to show for it are some fancy demos and trial deployments. Eventually you have to wonder what their future prospects are.
When cursor is selling for $60B, then grok has to really be worth several times that, right?
The average (and above-average?) investor really does not understand tech.
I'm not sure if you're following your own advice...?
The ~10% just sold was bought for $325 million.
The total price they paid was $1.205 billion ($880 million in 2021, $325 million now).
The $1.1 billion figure in the HN headline is kind of just wrong and presumably based on what they considered to be Boston Dynamics' total valuation in 2021, but represents neither what they paid for the ~10% nor the actual total they paid over both transactions.
Hey, Hyundai isn't "just a car company"
It's car manufacturing. Everything that could be done by a purpose specific robot arm bolted down to the factory floor is already done by a purpose specific robot arm bolted down to the factory floor.
What remains unautomated, then?
The long tail of tasks that are too minor, too finicky, too open-ended or too reliant on manual dexterity to be offloaded onto traditional robots.
This is where this new generation of robotics comes in. This is the kind of task they're designed to do: "a task that's still done by a human in a high automation environment". Universal robots are angling for the tasks that are impossible or uneconomical to automate with traditional industrial robots.
Hah! Hardly. I say this as someone whose first "real job" was in applying robotics research to automotive assembly - there are still a ton of assembly tasks that could be performed by a fixed-base robot arm, or a robot arm on a linear rail/fixed gantry. Wheeled mobile manipulators are only needed in a few cases, and humanoid form-factor is only "necessary" in very few cases (and I don't think the current crop of humanoids is particularly suited to these tasks).
In my opinion/experience, the impediments are that (1) the system integrators that are usually responsible for assembly-line robotics are too stupid to figure out how to apply robots to the problem, (2) the automakers themselves are often too short-sighted/stupid/unwilling to invest in increased automation (and particularly in building the in-house competency that they really need), (3) the hostile/exploitative relationship between (most) automakers and their main suppliers means that low-hanging improvements to parts/assemblies are a non-starter, and (4) the automaker C-suite (and investors) are too drawn to silver-bullet solutions (e.g. humanoids) than practical automation improvements.
Everyone in the industry learned that the hard way.
At a certain point, the tasks that remain stop being "dexterity" problems and start being "AI" problems. That is: a robot could do the task - if you either spent big $$$ on redesigning the entire task around the robot's intellectual limitations (uneconomical), or if you had an incredibly advanced AI capable of problem solving driving that robot (impossible with 00s AI).
The "universal robot" bet is the "incredibly advanced AI capable of problem solving" bet. That in 2020s, AI is finally capable. The body only has to be "good enough to make most tasks possible".
The auto industry is notorious for making incredibly myopic choices to save money/make money in the near term versus long-term investments. The relationship between automakers and their suppliers/vendors is basically a century-plus of the automakers trying to (1) outsource anything they can for a quick buck, and (2) grind the supplier/vendor margins down to nothing. (This is part of why the newer Chinese automakers with much greater vertical integration are such a threat to the traditional automakers; vertical integration has a high up-front investment but the payoff in flexibility and speed is significant).
The name of the reason is: corporate rot. They don't have the organizational backbone that wouldn't let their "in-house manufacturing" rot away into inefficiency and waste.
Not that it has much to do with why automation fails to penetrate certain tasks. The reason why "long tail" tasks are often beyond automation is: piss poor ROI, calculated correctly.
You go out of your way to automate a certain process with traditional robotics, and it'll probably pay off in 15 years. The chassis this applies to is going to be in manufacturing for 10 years. At least half the systems work you've done there would have to be redone for the next chassis. Fun.
The bean counters counted their beans, and found out that using traditional robotics there is a losing game. Thus the search for better options. And the humans performing the tasks in the meanwhile.
I actually don't think any of the big automakers have ever really, in-depth considered the ROI of "traditional" assembly automation (i.e. anything SoTA pre-2020), with experts in all parts of the process in same room. It's easy to assume that these companies must make careful measured decisions based on evidence, but in practice big decisions are made by small groups within the C-suite, often pretty divorced from the reality on the ground.
For example, many of the big asian automakers seem to have completely ignored the well-understood effects of their demographic crises (i.e. significantly aging population) on the future of their workforce (i.e. they are having trouble retaining and hiring new workers as the older generation retires) and this totally changes the economics of automation! Now they are all having to play catch-up, having realized that they must automate, at whatever the cost, because the issue is not "robots must be cheaper than human labor" it is "we might not be able to afford human labor at all".
An Amazon warehouse or Tesla factory tour would likely change your mind.
I had to do both of these in the last year and not a lot of humans around…
This is over the last decade at one of the largest automakers in the world. Naturally there is significant variation between individual lines and plants; some are newer and more automated, some are older and much less automated. Are some cars being built on more automated lines? Yes. But a great many, probably the vast majority, are being built with fairly low assembly* automation.
* There is a significant split in automation between "body weld" stages and "assembly" stages. Body weld is very heavily automated basically everywhere (although there are some surprising exceptions in places), while assembly is much less automated.
> I can’t imagine this would bring much actual experience with this new generation of robotics.
Luckily for you, my job has always been within the robotics research side of the company, so I am very much aware of the strengths and weaknesses of the current technology.
Indeed, Ford. Leaving aside the "old school" six axis robots that have been around for decades, Ford absolutely uses UR10s collaborating with humans to sand the entire car body in about 30 seconds, and to fit shock absorbers. They're also used at the engine plants. They also use the Symbio platform for transmission assembly, and fully autonomous forklift robots throughout their Tennessee plant.
The only thing I can think of are tasks that are so rarely done, it's not economical to build a robot for. But I then I also don't see how another robot solves this problem.
B) there’s a long tail of individual tasks it’s uneconomical to build purpose-built robots for each individual task. But it’s economical to have 1 robot that can do all of them.
The point is, human shape plus general purpose intelligence is an amazing combination to resolve the “long tail”.
Without the intelligence part, the body is useless.
Perhaps Boston Dynamics has that part resolved now too.
Do they? A human can both chuck kilograms of stuff across a room or kick in a door, but then pick up a single hair off the ground, or feel and manipulate (things even lighter than) a literal feather.
Robots can certainly do things more repeatably, if not more precisely.
But the rest of the human body is not useful in a factory environment, so the arms could be mounted on a mobile base that does not have any resemblance to a human.
One can lift insane weights, has insane torque, and absurd precision, and can do the same movement millions of times with virtually no deviation. You program these with an exact movement plan, just like you would programm a CnC with a tool path. They are basically cnc machines.
The other one is a inacurate, unstable, dynamic system controlled by neural networks and heuristics. It has massive deviation over each run, but that means that the programming must be able to account for it. Which makes it suitable to operate on problems that are messy, unrepeatable and human-shaped.
A robot that has to be carefully adapted and set up for the task vs a robot that you can point at a task and have it figure out how to do it. A robot that doesn't deviate vs a robot that absorbs all kinds of deviations.
It's a bet that The Bitter Lesson will win over Moravec's Paradox, in the end.
If I were Hyundai, I'd be looking at this as buying a significant amount of vision, dynamics, and integration systems expertise, not necessarily the dream of self-motive walking systems.
Better than what? It seems that as long as they perform the tasks "better" (cheaper / faster / lower-error) than the humans that are currently performing them, that is an improvement for the factory owner.
All you need to do is look at a recent video of car manufacturing process, and watch what the humans are doing.
>What remains unautomated, then?
Stuff that can be done by purpose specific robot arms on wheeled platforms, which is very difficult, but will be much more feasible than a humanoid robot doing anything.
Or a new take on car design with automated production in mind regarding all the wiring and what not (easier said than done, I'm sure many have tried and failed, but eventually someone will succeed).
This strike me more as a repeated internet myth more than anything else. There is near endless opportunity for purpose-specific robot forms.
And the tasks that change from day to day.
Even if I think this has legs, where do the cheap humans go to work after? Where/what are the remaining jobs for all of this displaced labour in both white collar and blue collar worlds? It basically screams UBI. And in a UBI world, the economy looks pretty different and humanoid esque robots start to look either very altruistic or very dystopian depending on how hard the oligarchs don't want things to change.
Like, what's the end game for humans in this path we're embarking?
https://xkcd.com/1053/
The comic strip is saying if above is true, then people still have to learn at some point so on average it would be around 10k people per day.
I think the math is this:
For people born in a given year: 4000000/365/30 = 365 people per day
but you have 30 sets of those people (those born this year, those born last year, those born two years ago, etc.) So 365 * 30 = 10950. 10k is easier to say for viral purposes.
https://www.bmwgroup.com/en/news/general/2024/humanoid-robot...
https://www.youtube.com/watch?v=CbHeh7qwils
Maybe what they're actually acquiring is Handle, not Atlas.
Perhaps they want to put some of the sensing and control features in, so a humanoid-like dexterity or adaptability for the business end of a floor-mounted robot arm?
https://www.google.com/search?q=nasa+robonaut+video+hand+why...
The gist of it is that all tools on the spacecraft (eg: space-drill, space-coffee-maker, space-airlock) are all designed to fit a gloved human astronaut hand. Waaaaay more complicated to make a robo-hand than a robo-suction cup or robo-claw, but then you are matching the environment, and guarantee tool compatibility against all extant tasks!
We already have specialized robots on earth... paper slicer, lawn mower, bazooka, whatever. They're all machines that are specialized for the task at hand, we're not making a humanoid robot that gets down on all fours and individually plucks blades of glass.
The car factories already have specialized robots... they're not mimicking a human hand holding a can of spray paint, shaking it up, and painting the car that way... it's a 6-axis arm, or a whole "grab the car and flip it while spraying paint" system.
It's not about inventing purpose-specific robots, it's about handling that long-tail of "stuff with tools that a human is designed to be able to use." Go over there, push that button. Go move that box from table1 to table2. Etc.
For well defined tasks in the factory domain, make a "real robot". For ad-hoc tasks in the interim... strap an LLM to a camera, battery, robo-legs and arms, cross your fingers, and hope for the best?
I think this is smart and not very risky. Tesla is playing a similar game with Optimus, for now Hyundai/Boston Dynamics is at least 5 years ahead.
The humanoid form factor is certainly may not be ideal but I guess they think the flexibility is worth it
They would be. When everything what could be done would be done by a robot. 24/7. Even without the lights on the floor.
I do think this might be tied to South Korea's demographics, by 2040 the working-age population is projected to decline 25% from 2020 and keep declining almost linearly until leveling out around 17M around 2065, a 50% drop total in < 50 years. I would think HMG / Hyundai sees a huge business opportunity or this might be a national-level political priority but I don't know the specifics.
While Atlas is the best humanoid robot so far, it still isn't useful in a car factory that's fully equipped with the latest factory robots that are strong enough to juggle car engines and are bolted to the floor.
Everyone knows the killer app for humanoid robots is building the Mars colony amirte?
My personal take is that one of the reasons is their posture against ML. They've been very "GOFCT" and have only recently started to incorporate ML concepts.
Well...there is the uncanny similarity to the T-800 and and uneasy realization that the owner of BD could become Cyberdyne Systems IRL. Perhaps some companies like that notoriety but not sure if many want that.
https://terminator.fandom.com/wiki/T-800
This seems like a small correction if they wanted to reacquire and clearly the market isn't valuing BD all that high.
Why do you think it's one of Sundar's biggest mistake?
That seems out of character
All those tokens have to go somewhere
Who cares about quality. Speed is the new black.
On a relevant note: https://www.theverge.com/news/674366/nick-clegg-uk-ai-artist...
Start to be sloppy somewhere, you'll be sloppy everywhere. As we "learn and enable" to do things faster with less effort, the quality of the thing we (as in humans collectively) do decline.
AI, when used as the sole blunt instrument, accelerates this dramatically.
Source: https://www.kbb.com/car-advice/car-warranty-guide/
Teslas new car warranty transfers as is to the new owner.
Source: https://www.tesla.com/sites/default/files/blog_attachments/m...
Seems like the effect would be to take off a significant chunk of value off your used car in a way that makes it more difficult to buy a new car.
There's no shame in being broke, of course-it is merely a catastrophe. The fourth owner of a '11 Sonata is gonna have a different outlook than the fourth owner of a '73 Mercedes 600 Pullman.
the other timing here is the increasingly belligerent union who are demanding pretty outrageous compensation for what a typical American worker would make. I think the goal is to immediately automate the workforce and move the plants out of korea speaking to insiders.
Rest is previously reported stuff.
Related from earlier in the year:
Hyundai Introduces Its Next-Gen Atlas Robot at CES 2026 [video]
https://news.ycombinator.com/item?id=46520508
And old discussions when the intial buy happened:
2020: Hyundai to acquire Boston Dynamics
https://news.ycombinator.com/item?id=25363981
2021: Hyundai acquires controlling stake in Boston Dynamics for $880M
https://news.ycombinator.com/item?id=27588047
However, let's hope they will keep on doing cool stuff under their new owner.
Car factories seem to be a pretty good initial market, given that Tesla is doing Optimus and Figure has humanoids in a BMW factory. But the whole point is that these are general purpose robots, and there are lots of other factories. By the time that market is saturated they'll be capable of more.