I'm not sure why some comments here are so pessimistic. I was grinning ear to ear while seeing the video. You're telling me we've finally got robots that can do our laundry? And the model is at least as free as beer?
To quote the lyrics/meme, we used to pray for times like this. This is what robotics should be used for.
They are likely Effective Altruists highly attuned to AGI (With current LLM's) risks, fearing an LLM uprising and human extinction, which they regard as the world's most critical challenge.
the name stopped being relevant a long time ago when they went from rating charities to coming up with horror stories about ai. "ineffective egotism" would be better for what they are now.
I've definitely noticed even more than that - any submissions painting the current USA in a good light are instantly and mercilessly buried (you can check my submissions for examples), while anything even remotely on the "USA bad" side gets to the top. It's really a sad state of affairs.
You may take this as “USA bad”, comment, but let’s take the U.S. life expectancy reaches record high (in 2024) link you posted.
That sounds like a good thing and it is. But it’s not a US story. In fact, other wealthy countries didn’t just already have a higher baseline life expectancy pre-COVID, they rebounded to it post-COVID sooner than the U.S. was able to its already lower level.
So what this actually is, is a story of “global modern medicine good”, and unfortunately, even within that, the U.S. is among the worst performing.
I tried to read the Washington Post op-ed piece you shared on reasons to be optimistic about America’s future. As soon as I scrolled down, the content was replaced by a patriotic fireworks animation and a paywall. Seemed a bit on the nose.
aka this very thread - top comment and most of the entire comment section has been taken over by literally "how dare anyone say anything even remotely negative about China, what's wrong with you".
Is that so shocking on dominantly US forum? I wouldn't expect different situation, people are not 100% objective even if they try hard.
That being said, >95% of the world lives outside US and in recent times (culminating with orange man's second term but started way before) US has alienated literally every single other country, democratic or not. This is what happens when reins are firmly held by unstable people full of pettiness and mental issues, plus visibly corrupt too.
So division deepens as seen also here, and as much as 20th century was a century of US, 21st century is and will be one of China. Every move US makes now just pushes whole situation further into that direction. So folks cheer for their own camp, not some larger common group of aligned friends.
It's strange because my impression has been pro-Chinese comments on HN. I suppose it depends on the topic, the time of day, and how many bots each government is employing on a given day for the key words and their compute budget. This submission seams rather balanced to me.
Is there evidence of any state-sponsored bots at all? It seems like this has become an accepted fact but I don’t see any data indicating it’s actually real.
HN has lots of hellbanned bots lately (I have showdead on [1]) but I never notice ones pushing a political slant. But I suppose nation states conducting a campaign on Hacker News would be much too cunning to make the bots obvious
[1] because many posts get wrongly flagged, apparently in part by trigger-happy users who will flag anything with incidental similarities to AI prose (everyone, PLEASE STOP DOING THAT), in part because HN's automoderation is severe for new users, and in part because plenty of people get hellbanned either unfairly or although they do still write plenty of worthwhile posts. I even rescued someone who it turned out had been hellbanned by a mod by accident.
Well. I already decided some time ago that I don't want to have my
money go into US corporations as long as the orange King is living out
his genocidal fantasies (and even without that, I am already annoyed to no ends about Apple, Microsoft, Facebook and Google abusing everyone else so blatantly and violently, including buying legislation - look at the mandatory age sniffing movement before you can access information on the www any longer; this is fascism, and children are abused as fake reason for change here). But I am not all that happy about China becoming
so dominant in the whole product chain either. I think the production
should be more balanced world-wide, globally. And also, very much so,
locally. I am also aware that this is very hard to do, for various reasons,
but still if I have a little influence then I really would rather not
empower the USA or China any more than they already abuse everyone else
here. And both do, by the way.
Right, but the original commenter said "living out his genocidal fantasies", which means he is actively performing genocide on someone and it also implies this is something being done with active, excitement.
I'm simply asking who are these people being genocided?
western elites are currently at odds with the chinese elites and every impressionable idiot is being inundates with antichinese propaganda.
Lets be frank, every single issue we have with the CCP is either also being done by the west or in the works.
chinese oppression of the uighurs is nothing compared to whats going on in Gaza.
china forces prisoners to peel garlic while american prisoners are paid pennies a day to fight wildfires while being prohibited from being paid for those very same skills as firefighters when rehabilitated.
america is quickly workin on legislation to to make criticism of certain foreign governments illegal with very real reprucussons (revoking passports).
At least china has a working power grid and high speed train network.
It's possible to intelligently discuss the merits of Chinese society and also acknowledge that it's a brutal autocracy.
Making some kind of equivalence of the two in terms of the rights of the individual is so ridiculous that nobody is going to take anything you say after that seriously, you do your arguments a disservice by doing it.
"Yet" is the keyword. it won't be long before we arent' any different from china at the rate we are going at and I think a lot of us would be better off preventing the west from turning into an oppressive autocracy instead of judging china.
It's also relevant that China isn't open sourcing results because that's something China will do forever on some philosophical grounds. They do it because it is currently in the best interests of the Chinese state (and the world, as it turns out, coincidentally) but as soon as that stops being true there is zero chance it'll continue to be open source.
Yeah the US needs to step up with its Open AI research here, they should start some kind of "Open AI" collective to release models free to the world, do it all open like
Is there someone saying you should hate something?
If we want open source technology we should recognize that the current arrangement is non sustainable - the moment china doesn't have an american AI industry foil to play against it'll immediately close. It's up to the US to build more open source models. We already own the content used to train them, collectively. We should own the results as well. Make the big AI companies release weights.
Eh, I take what I can get. A lawyer doesn't argue that his client is innocent (though that might be tha case), because of some innate sense of justice, but because he gets to paid so.
This super interesting link showing things that were impossible even a couple of years back, sitting below the day old news about Kagi's me too browser tells everything about HN audiences. West will fall behind in robotics and mostly because of the wrong priorities of its VCs and developers.
The singularity is literally here right now! Some people will never be happy! This is easily the most amazing thing I have seen in my life! These things will be taking care of and raising babies very soon!
> You're telling me we've finally got robots that can do our laundry?
No, we don't.
I wish we did, that would be amazing.
There have been demos like this over a decade. But they are just that, demos. Demos done in a controlled environment.
If there was a robot I could buy that could effectively fold laundry or do dishes, I would gladly drop 50-100k on it, but such a thing doesn't exist and likely won't for a long time.
You would honestly pay 50-100k on a robot that just folds laundry or does your dishes? In the US you can get a maid that cleans once a week for $150-250 and would do a much better job. By the time that cost would equal this robot, the robot would be obsolete, not to mention any maintenance costs.
Also dishes are the easiest chore to do around the house, especially if you clean them after use instead of letting them pile up.
I honestly think I would. I have a maid come every two weeks which is of course not often enough for dishes/laundry. A maid that only came weekly couldn't stay on top of dishes either but they could do laundry fine sure. I do a lot of cooking and I find constantly doing dishes and loading/unloading the dishwasher to be a huge never-ending pain. A device that costs 100k and would free of me that sisyphean task forever would be a godsend. Also you can have the robot work anytime whereas a maid is probably on a schedule.
And not to move the goal posts, because my original post did only mention laundry and dishes, but I assume once a robot reached a point where it could do laundry/dishes effectively without supervision, that it could do a variety of other tedious household tasks. 50-100k would be well worth it IMO.
Yes but unemployment where a small cadre of rich elites own the robots and the rest of us are serfs is very different from the public owning and benefiting from the robots in the form of strong social services and UBI.
Those of us who liked the open web and P2P and hacking on projects all the time are seeing these models as a way to rebuild the internet in a shape that pleases us. To build anything and everything we want. Sans opportunity cost.
And that's the real value - you don't have to burn the candle of your life away to get what you want. You'll have so much more time now that you aren't AST wrangling.
Opportunity cost is vanishing (or at least the cost is multiple orders of magnitude smaller). And that's the most exciting thing that's happened in tech since the internet.
I'll probably be able to make my own phone operating system.
I'll probably be able to decompile N64 games, port them to Rust, combine them, extend them. In a weekend.
I'll probably be able to make my own music management platform while simultaneously working on a few robotics projects.
The whole thing feels like that the US collectively isn't feeling confident about itself in the coming creativity-limited world that AI/AGIs supposedly bring, and that it's not happy about that pessimistic rational projections. I don't know exactly why wouldn't it be or why nothing seem to be done about it, though. Both look to be fundamentally solvable problems to me.
> You'll have so much more time now that you aren't AST wrangling.
So much time because you were fired and have nothing to do. Nothing is as good as oodles of quality time worrying about funding your family's needs. It's amazing.
Robots will take jobs from massive amount of low income people.
AI will take jobs from a massive amount of avg and high income people.
The disruption ahead is getting pillowed right now due to physical limitations of progress like amount of energy, building DC, doing research, training models (its still takes 1-3 month).
Also a massive buffer will be the boomer going into retirement.
Depending on the speed of all of this, we might get a lot poorer before we get abundance everyerwer
But there's no feeling of power with giving everyone bread. Billionaires seem to have more than just an addiction to big numbers, maybe seem to want to rule over others.
Also: you can literally make incredible software that would take half a year right now for $200/month. Or free if you run local open weights.
You can do branding, marketing, design. You can create comics and movies and video games.
You can make an AI tool right now and get customers today. You have no idea how wildly popular this segment is. How much people willingly engage with this stuff and want it.
Yet you choose to disengage and complain about a future that hasn't happened.
It's more likely that firms will need to hire more to stay competitive. Or that workers will fire their old jobs and break out on their own.
Something I used to think was impossible: competing with Google, Apple, Stripe. Not anymore. I think you can compete with anyone.
I saw robots doing this in tech demos 15 years ago. And even those robots weren't doing anything new. Like all AI models, this is just AI repackaging and rebranding of things that robots have already been programmed to do using more classical control methods. The idea to do everything via AI was of course thought of for a long time, but the conclusion has been we have faster and more reliable methods to do the same thing with less compute, and that's still true.
During my undergrad I worked on a project where we set off a bunch of walking robots, recorded their walks, and then used the data as the basis for a learned walking controller. I see this Xiaomi model as pretty much like that.
It's a nice technology and maybe could lead to better folding one day but unfortunately it's not "finally we have robots that can do our laundry" (yet).
Yeah exactly. I've been seeing demos like this for ages. I'm kinda shocked at the lack of skepticism in the comments here. The hackernews comment sections have changed a lot these past couple years...
Is a Chinese robot vacuuming and mopping my floor every day? Yes it is. Could I do it better manually? Absolutely.
I still prefer the robot to do an ok job than to do it myself.
Most people in software misunderstand how physical goods work. When making machines you don't care about getting everything perfect all the time, you care about staying within tolerable limits all the time.
How bad do the folds in your laundry need to be for them to be no different then just being dumped in a pile? As long as we are above that point the robot is useful.
I think there is a stark contrast between the plus or minus micron level tolerances, and whatever is being churned out of LLM ('not getting it perfect all the time')
Eh, this still has value for me if it can do it reliably. Leaving a robot to badly fold my t-shirts overnight and then badly put them in my closet the next day is fine. I don't need my casual wear beautifully folded. But I do need it organised, and that's apparently a chore I'll just never learn to like doing.
Physics. Right now the clothes folding is computationally limited, but assuming we get faster computers to get them folding faster, gravity and air resistance are still going to take their time.
Exactly. I will pay money for a robot to take laundry out of the washing machine and hang it to dry and then fold it and put it away or just sort it by size. This is a chore that my wife and I have to do several times every week.
My solution: stuff the clothes from the washer into the dryer, then stuff them into my wardrobe. I buy clothes that don't get wrinkled much even if simply shoved in the wardrobe and if they do, I tolerate it and wear them anyway.
Ha! You remind me of something I used to tell my colleagues back when I held a regular job. They would compliment me on my highly patterned shirts (eg. A very small floral pattern repeated). I would explain that a shirt like that just requires low spin in the washer, and then drying on a hanger. Whatever wrinkles are left are almost invisible due to the pattern. With collars and cuffs I’d just literally manually stretch/tension them while they were still wet. Bada bing bada boom.
This is a mostly American perspective that I find endlessly amusing. I only have a handful of clothing items that say that tumble drying is fine on the tag, and most of my clothing is very casual wear (t-shirts, jeans, sports clothing). Dryers eat up clothing like it's nobody's business. Plus, electricity costs money while the sun is free. And while yes, the time to hang laundry has value, it takes 10 minutes to hang a load. If it rains, I hang the clothes indoors with a dehumidifier which is cheaper to run than a dryer and dries clothes in a matter of a couple of hours.
I was referring to shirts specifically. If you put any kind of cotton shirt in a dryer it’s almost certainly going to need ironing. Waiting for a shirt to dry naturally shouldn’t be a big deal.
So what? it can take hours to do it and I can focus on something else. If these robots actually become available on affordable price I would buy one without blinking. It would totally free up my weekends, no more hours wasted doing laundry, cleaning the apartment etc.
Sure but that holds true for most new tech. Expensive and slow but it can only get better or stay the same. I'm thinking it's only going to get better.
> I'm not sure why some comments here are so pessimistic.
Actually, it seems like 99% of the comments are glazing this with zero skepticism actually. If anything the comments are too enthusiastic and lack the measured discernment that was more common on HN 5+ years ago.
I disagree. We should use robotics to extend our reach into hazardous environments, enhance our physical capabilities, or to provide a steady handling of things where it's absolutely needed - doing laundry in my household is none of these.
Folding laundry takes me 10 mins every 10 days. I organize my closet once in a while according to my sense of fashion, and now I'll just need to keep that structure. I don't always like to do these chores, but I have no issue in acknowledging them as part of my responsibilities. Why would I even automate[0] that?
Robotics is costly, and I see no way how that would change anytime soon. That isn't a flaw of some kind either, it's rather just one of the limitations that needs to be accounted for, neglecting that would be detrimental to the future I envision.
Don't get me wrong, I love robotics with a passion, as engineer I like to tinker, and I'm a big admirer of FESTOs bionic research[1]. It's just that the field suffers from an "overpromised and underdelivered" aura as far as I can remember, and I remember the presentation of ASIMO[2] 26 years ago.
Some people would like to see the field hyped up, focused on the consumer like every other field in tech, probably for financial gain, but I personally see the value in its downplay. To be honest, I love these reality checks, as I see them as a valuable part of the learning process all together - a journey for what it means to automate, and what it means to define as human natural chore.
Sure but the robots that extend our capabilities often start by replicating our current capabilities, even if those are as simple as folding clothes. It provides a baseline for the next generation of robots that can do more.
You could replace "robotics" with "AI" in what you wrote above, and it would sound like what people were saying about llms when they first came out. Not trying to glaze one way or another but historically this take is outdated.
A subset of the people on HN hate AI and yell "slop" at everything they see. It's even worse if you go to other social websites.
Just ignore these people. If they don't use AI, they likely won't remain practicing in our field for much longer. They will naturally self-select themselves out of relevancy.
I was Claude coding and got hit by the 3 AM reset bug. I went on reddit and saw hundreds of other engineers hit by the same. We are different people than the AI haters, and there has never been something so powerful it's kept me up past 3 AM every single night working beyond every fiber of my body telling me to sleep. I feel empowered as if I command an army. And there are millions of us. The sheer might of all of this will win out over negativity.
I've built multiple successful businesses in the last year. If you're doing this with any sort of interest in self-employment you don't need your employer anymore.
I have never been happier. This is the best future, and I feel lucky to be living it. It's what I dreamed about when I was a kid, and it feels magical. As far as I can tell, the future has no limits.
It's always funny seeing people that write like this as if they are somehow exempt from the future we're seemingly rushing into. Bro, you prooompting yourself into unconsciousness is not going to make you not obsolete.
Unless you're being sarcastic in which case poe's law.
Working past your personal bedtime, part of an imaginary movement of millions might, ignoring the negative aspects of AI, empowered like commanding an army. This frankly sound psychotic, and it's exactly why uncontrolled AI usage is seen so critical by many.
And all other devs will leave the field just because they prefer to work differently? That is not only not realistic, it is violence fuel.
The frustrating part is that these posts will never post any details about the business, despite the fact that it would immediately invalidate most criticism.
It seems to be a similar phenomenon to productivity "gurus" who write and evangelize polyphasic sleep and "grindset" and so on - when you ask what they're using these superpowers to actually accomplish, they're very quiet. Or it's "coming soon." Or "I'm not going to waste time explaining to someone who obviously doesn't ~believe~." It quickly becomes religious.
> there has never been something so powerful it's kept me up past 3 AM every single night working beyond every fiber of my body telling me to sleep. I feel empowered as if I command an army. And there are millions of us
Yelling "slop" at a clothes-folding robot is stupid, but reading this kind of shit makes me want to embody Ted Kaczynszky and go live in the woods.
Ahah you AI bros can't stop talking about Claude and llm, and telling stories of how great it was that one time, even when they are not the topic at hand.
> I've built multiple successful businesses in the last year. If you're doing this right you don't need your employer anymore.
We nearly destroyed our planet to get there. Technocrats here applaude the new technological achievements that make life a little easier (less boring?) for the top one percent while willfully ignoring the associated costs for everyone else. Manufacturing these at scale will only make things worse. We are approaching peak decadence ever faster.
I am of course impressed by the model and the robot. Great stuff,
But the downside is that won't we all just turn into the obese, stupid humans at the start of the WALL-E film?
Mindfulness teaches us that we can enjoy the washing up, the laundry, planting seeds, tending the garden. It connects us to something. Without this will we end up with a rise in mental health issues, because for one thing people will just end up on social media in the time they save.
Ironically, the people who can afford to have in-house help to handle things like cooking and folding clothes - the rich - also tend to be fitter and more active
Maybe if you can shave an hour every day from chores, you will actually have the energy to enjoy physical activities
I often wonder if they just made us do more washing, use more cups, travel further in the same time and get more expectations from our boss though.
I've got 2 kids. I'm always doing laundry and loading the dishwasher. How much of that is just higher expectations of cleanliness? Back when washing clothes involved going to the river all day, kids would get sewn into their clothes for winter. These days they have 5 outfit changes a day. I'm certain that my dishwasher means everyone just grabs a clean mug/glass/bowl rather than just rinsing a dirty one. I use a computer all day to get stuff done. But that just means my boss expects more stuff to be done. The baseline shifted.
A few hundred years ago people mostly lived near where they were born and didn't travel much. Now we get in our cars to drive a few hours to visit our family and friends. But they only moved far away because the car enabled it.
Obviously there are benefits - especially for those who find dirt upsetting - but have we actually changed the balance of leisure time very much?
This discourse reminds me of the handful of Christians who thought we needed Corn Flakes and uncomfortable clothing to stay chaste and that electricity made us stray from God.
Chill. Your brain is the product of the past. Reread that: your brain is anachronistic and outmoded.
We get to live in the future now. Count yourself lucky. You're one of the few that got to witness the birth of flight culminating in landing on the moon, except in this case your floppy disk became your personal shopper and stay-at-home Michelin star chef and part time driver.
It's okay to feel a bit of discomfort. It's hard to believe this is actually happening in our lifetimes, but it is! And that's so cool.
For a bit of context there are some classically very hard things going on in this video, so if you're not impressed, be glad you didn't have to try to build this 10 years ago.
- Coordinating two hands
- Mobile body
- Deformable objects
- Thin affordances (bag zip in particular)
- Multi-object single grasp
Really a host of things that had individually been tackled as either perception or manipulation problems and probably led to more than one person's PhD here on HN.
Watching the two-handed robot has me thinking that a third hand, perhaps with a different set of capabilities (like a tentacle grip) might make many tasks easier. I realise that more limbs = more problems, but even so, having additional ways to hold and turn things would be a big help.
My thoughts exactly, but to train it you'd need a lot of data with someone manually operating it. With just two hands it's enough to have a lot of footage of humans doing it. Maybe down the line the robots will be able to self-improve or someone will come up with ways to optimize this..
SciFi novel "The Mote in God's Eye." has aliens with three hands.
One "gripping hand" kinda coming out the gut, and two dextrous hands for detail work. Led to abuse of English language with:
On the one hand...
On the other hand...
On the gripping hand...
Which is reall just a patch for the problem of "On one hand… On the other hand… …On the OTHER hand…".
It's worth noting that the one-other-gripping pattern seems meant to convey two considerations and end with the strongest of three.
The English one-other pattern doesn't necessarily require the second consideration to dominate. They could be equal, or they might merely introduce concerns that haven't been evaluated to conclusion yet.
Alas, one-other-gripping doesn't seem to generalise to more than three concerns.
I feel like there'll soon be a market for AI controlled enhancements like limbs. Neural interfaces seem to be a long shot, at least for now, so some kind of AI that senses the environment and intelligently react to it seems like the next best thing
When you Robot folds your clothes but it's kinda wrinkly and sloppy but you accept it anyway because it's easier than precisely folded and pressed clothes.
Isn't that the story of <any> technology? When the coffee machines first came to be, Barista was definitely a better option. However, through the evolution, in many cases today machine does better job than barista.
Better analogy would be that even though the coffee machine is still worse, we choose it because it's more convenient (at least I think that's the case).
Anecdotal, but the best machine will never be able to do something as good as the best human in most circumstances. That is because, unlike machines, humans have passions and character that subtly imbues itself in the end result. This, of course, isn't limited to coffee, but to anything a machine could do that isn't strictly tied to exercises in logic (for example, computers will always be better than humans at most games because they can be analysed logically and acted upon)
I feel like, despite not being my (and presumably most of HNs, given the general demographics) cup of tea, sports are largely that. The players are skilled but also uniquely themselves. The world cup wouldn't be nearly as interesting if it were just unthinking robots fighting to see who can run around and kick things the best
I'm not sure what the :/ implies, but if you're disappointed by your own folding techniques, spend a bit of time on it, look up some guides, and fold clothes with purpose. or something.
(Rise up against the machines! Folding your own clothes is an act of resistance!)
I've always taken :/ to be a kind of hopeless dejection mixed with apathy. In this case, it seems to convey their disappointment in their skills as well as the futility of improving
I don't know why you would assume privilege from a comment like that. If anything, it's the unprivileged kids who would know that you'd get your hand slapped for folding clothes like that.
Speaking as a relatively unprivileged kid from Southeast Asia, my parents were so busy working they just didn't have time and energy to care about me, let alone how well i fold my laundry.
When AI is a the forefront of scientific breakthroughs and all human endeavor is painfully minuscule compared to AI, we'll still be calling it slop as we all hypocritically generate more and more slop.
Nobody is proud of this. Where the hell did you get that idea?
I think weak and scared people tend to not want to face the truth about AI. I said the above not because I support it (I don't), I said it because it's the god awful truth. All trendlines point to steadily improving AI, and tons and tons of people continue to call it slop even though it is clearly becoming better.
My question for you is, in order to sleep at night do you have to lie to yourself? Do you make up stories about how AI will never be as good as human intelligence and do you assume anyone who says the contrary is an ardent supporter of AI? Have you thought that... maybe nobody really supports AI, they're just telling the truth?
Weird my ass, just stupid people who clearly assume the wrong thing.
this is what excites me most about the probable AI future. househelp is cheaply and easily accessible for most people in the east, however being able to afford a cleaner or maid in the west is a luxury. i spent a ginormous amounts of time every week keeping my house in shape. i look forward to the day when i will be able to carve out that time and get it back for myself.
Househelp isn't cheaply and easily accessible for most people in the east, a large portion of people in the east is just poor but you don't consider them in 'most people'.
Of course, you can mention all the things they can't do (what if there are stairs, what if the doors are smaller than the robot, etc) but it's mindblowing to see what they can do - and also the chat panel is really nice and makes it more interactive.
Designing a non humanoid robot is the way to go. The design of new tools and appliances for hybrid human/robot use will eventually follow: the zip line for the shoe bag required way too much precision work.
I think this perspective is now demonstrably false. It was basically the defacto position of most roboticists until a few months back. But we've already seen the cost of humanoid robots fall precipitously (even without the kind of mass production that genuinely useful home and real world intelligence would bring).
Which seems more likely, that cheap humanoids will continue to improve in dexterity, battery life, strength etc. Leveraging the vast investment in research into real world task training. Or - that we'll redesign millions of everyday objects to fit the limitations of current task specific robots?
If you answer the latter, I've got a whole bunch of mid century quirky labour saving inventions in the back room I'd like to sell you.
I can't believe someone shares my point of view. We have already many non humanoid robots to do boring stuff (underwater maintenance, stuff like that). For some reasons people tend to think that for robots to do stuff close to humans, we need human like robots.
We do not. I want a swarm of insects-like robots to fight mosquitos. A snake one that fixes my pipe. A spider-like robot that cleans surfaces. Etc.
I will suggest mix of both: some everyday tools are already built for being used with hands and legs (e.g opening a door). But indeed adding some different grip types, additional arms, etc, could be helpful.
This is amazing! I can see in a few years having Home Assistant use an LLM to coordinate my home while I'm away. Tell the LLM to make an inventory of my home and then send the vacuum and the robot around to clean everything up. A multimodal LLM can inspect pictures for items that need to go in the dishwasher, the laundry or needs polishing and reorganise my closets to best allocate space, make the beds, take care of the plants. It has all day to do it while the solar panels provide ample energy anyway.
As a house spouse I want this! I _hate_ folding laundry and always procrastinate on it. Doesn't even need to be able to load the machine like they additionally show, "just" folding it is enough for me!
Does this require a given hardware platform, or for example, could I run this model with a cheap mini robot with a single camera (i.e. good enough without depth), tracked treads, and an arm?
When the page says "embodiment-free (UMI)" it means they've used a standardised gripper and camera arrangement [1], while aiming to be agnostic to the type of robot arm that gripper is attached to.
I have Roborock (Xiaomi) automatic vacuum and it's fantastic technology. iRobot had a 10 year lead and did nothing with it. The robot is competent at avoiding clutter, clean the mops itself, leave my floor sparkling and so far has been extremely reliable.
Got two Xiaomi phones in a row, one in 2018 for 250 euros (note 8 pro) and one in 2023 for 300 (13T) and I liked them a lot.
My gripe with the second one is lowish battery life, it always bites me on longer trips when my gf's Pixel 8 lasts considerably more, but was also way more expensive.
Will check battery reviews better next time I shop for a phone, I don't care much for the rest.
I just can't deal with all the ads on Xiaomi phones. Switched over from Chinese phones, still remember my JiaYu G4 fondly, to Samsung few years ago where the situation is a bit more bearable
Anyone have any recommendations for robot hardware where I could test this open-source model? It would be a cool/fun project to get it working on a little robot.
This is very cool, and the paradigm that scales to bringing truly useful robots to virtually all domains of life - just as LLMs have scaled to that degree across virtually all domains of knowledge work.
I picture a lot of perfectly tended gardens growing food.
Please help me understand. Is it more effective to have a general model like this instead of a state machine that leverages other models and a deterministic set of code?
How are these models compared to world models by Waymo and Tesla, assuming their models are trained for recognising objects that can be further generalized for everyday chores?
Compared how? Architecturally? In training methods? In performance?
The remarkable thing about this modern "full VLA" approach is that it doesn't train for an explicit "world model" at all. There's no factored world state, no pool of trajectory rollouts being sampled from. Instead, everything is rolled into the same VLA stack that does all of it internally.
The world model exists, but it's implicit - the VLA learns one, but only because it helps its training objectives. Trajectories are planned ahead sometimes, also implicitly, for the same reasons.
It's a very "bitter lesson" way of doing things.
As far as I know, Tesla's stack is somewhat similar architecturally - a highly integrated end to end model - but the training process for FSD does include factored "world modeling" side objectives to force the stack to develop driving-relevant internal world representations. So it's in a middle ground between a classic "world model" system and this flavor of VLA - albeit surprisingly close to the latter.
Whether the world modeling side objectives are principled and justified or a leftover from the days of insufficient scale is unclear.
From even before the RAM shortage there was noise in the robotics world of sensor and actuator shortages due to them being bought up in bulk by some of the major AI labs.
The first few second of the shirt folding arms trying to put the t-shirt properly on the table to start folding it are at the same time amusing and scaring.
What if the thing retrieves some anger-driven post from Reddit from its training data set and start literally to flip the table??
Neither the Github nor the HF repos have anything on the XiaomiRobotics-1 model or datasets :-( They have a bunch of assets (datasets, code, model, etc.) for XiaomiRobotics-0 however, but I don't know much about that family of models.
Yes, it's truly remarkable how small modern VLAs are.
Even Tesla's FSD AI stack is small by modern LLM standards. Once the hardware is there to enable real time 100B inference on endpoint SBCs? We'd see gains in AI robotics from that alone.
WAIC has been wild this year. We got Kimi K3 and Qwen 3.8 announced. We got robots boxing and open-sourced robotics foundation models. And we got the announcement of WAICO.
At what point are we gonna start seeing American firms attending WAIC?
In the domain of hardware, what is the equivalent of a harness? How well would this generalise to any robot, versus one that is specifically built the way Xiaomi did?
Considering that they pre-trained on non-robot data, and mid-trained on data from other robot types? I suspect that generalization would be decent for most similar platforms - things with 1 to 2 arms and 2-finger hands.
You'd probably have to fine tune for good performance on a specific arm type. But this model isn't trained against hard assumptions of what the underlying hardware is.
How are there still people who look at all this advancement and the only thing they can utter is "it's not there yet?" It's clearly here, and if it's not to these nerds' expectations, it'll prob meet it in 4-6 months. This was cool asf.
Because you see a robot do one thing, doesn't mean it can do other things. We've had good robot demos for decades.
For all we know they could've spent 5 years and 10 million dollars on getting this robot to fold laundry. That doesn't mean it'll be able to mop the floor better than a Roomba in a decade.
No we haven't. Back in the 90's Honda's Asimov could barely walk. Now we have robots managing to run marathons. Clothes folding had to be played back at 50x speed because the compute just wasn't there yet. It still is barely there, some of those videos are sped up. They've probably spent way more than 5 years and 10 million dollars developing that robot. Roomba already came out with a mopping robot, why would you want this one to do that?
It's probably because robots have been doing cool asf demos for half a century and they're still mostly just demos. Elon has made a lot of money greasing the 4-6 months away crowd.
They're trying to show their investors a butler robot that can do all types of household chores.
Folding fabric has traditionally been very hard to solve with robots; the ideal places to grab the fabric might be hidden, the fabric patterns are repetitive, and the material flops around, tangles in the grippers, and brings other pieces of laundry with it.
Successfully sorting and folding a basket of laundry shows they've got a control system capable of performing some quite difficult tasks.
They also use a large ring attached to the first zipper to help the robot grab it. This is not a realistic zipper, but I do appreciate them being honest and showing that grabbing small objects is still not perfect
All you need to do to understand what task is more complex (for a robot to solve) is just to ask a 6-8 years old child to do it. It takes days before they "understand".
So it cannot be that simple (for the reasons you mentioned).
As a parent with 3 kids and both my wife and I working, folding clothes is the most tedious household chore. It's physically and mentally exhausting (unlike doing the dishes, which is only physically exhausting as I can wash piles of dishes while zoning out). And not to mention take so much time, a week's worth of unfolded pile of clothes for our entire household can take me 1-2 hours to fold uninterrupted (2-3 hours if I do it while watching Netflix or something).
Small sharing tips, in case it can help: have the maximum of hangers possible. It take a bit more space, but the time saved is huge + it remove iron time for low wrinkle sensitive clothes (e.g cotton polo)
Yep.. and with 3 kids so many of the time-consuming "tasks" are not open to automation because a human doing it is the point.. eg. bedtime books, bathing the kids, helping them make their lunch/breakfast, etc.
FYI I do have 3 kids, no, it doesn't take that long (we share this with the wife), in a few years the kids will be able to help us with this task so it's gonna be even less time. Doing household chores is good for the kids. I wouldn't wan't this robot driving around my household when I have small children anyway...
Don't you just throw all the washing into the washing machine, wait for it to actually do the work, and then hang it all up afterwards? It's like a 15 minute job max.
I mean folding after it has been dried. I hate folding clothes. In my case luckily my wife doesn't mind (as I don't mind doing the rest of the process) but if in say 25 years from now there will be a relatively small and affordable folding appliance, I would totally buy it.
I don't know, was already bitten by this in ~2015 with self-driving cars and thinking that my kids will never ever need a driving license so now I'm really cautious on these topics.
Yes, but anything outside L4 or L5 will still need a driving license to drive the car. In 2015 I though that around 2035 that would be the case, but now I doubt it seriously.
80% will be in the laundry bin, your first task before doing anything is either shouting around the house to bring in the other 80% right now or go on expedition yourself.
10 minutes later you are at the washing machine, where you'll never be able to fit all laundry so it's two runs minimum. Somebody will hopefully in advance, but probably only prompted by your call around the house mention a number of particular items will be needed later today, certainly before the second load will be dry. You are now searching for those items. Or arguing for why you really don't have time and parents are humans too and not unpaid servants.
Another 10 minutes later the first load is washing. You set a timer, or perhaps have a machine that pings you when done. Of course, you have More Shit To Do, so you don't notice your alarm or ping and by the end of the morning when you hoped to sit down to enjoy maybe a whole 15 minutes of reading move the first load to the dryer and put the second load on. You discover that somebody never emptied the dryer. Finders keepers!
1,5hr later you fortunately do not miss the alarm and take our the dry but not 100% dry laundry and move the second load to dry.
3hr later you empty the second dryer load, because the run took longer because nobody ever fucking cleans the lint-container. At least you had time to fold (which to the many item-types in your household and your poor fine motor control certainly does not take you _just_ 15 minutes). A cat decides to kick over a pile and leave hair in a second pile, causing you to redo some work. 30 minutes in your are done and finally have time to fold the second driers worth.
Everybody's different, I guess, but for me this is absolutely a chore I would automate away, and I don't want to "force" another human being to do it for me.
A lot of smaller chores look trivial when considered in isolation, but have disproportionate time and mental cost because of how they fit in the flow of life. Sometimes the effect is similar to interrupting an engineer for a couple of minutes.
It is important to remember that China is not only about open weight LLMs, fastest cars and flying taxis. There is still a lot of very serious human right abuse going on in this country.
I'm not super sure how relevant this is to the overall topic and thread, tbh. There's plenty of cheap yet real rebuttals, whether it's all the things with ICE (and not just the top-of-mind stuff; but effective deprivation of due process for suspects; etc), or even war crimes.
I don't think it serves much purpose tbh -- it's not going to change anyone's mind.
Yeah, western lecture on human rights carry no weight these days lol (Israel support, Trump doing Trump thing, war in Iran, etc). They ruined their own reputation so the rest of the world just tuned out.
When the last day's news were about other people shot dead by ICE without a plausible reason, it is hard to take seriously discussions about human rights abuse coming from USA. It is hard to imagine a more serious abuse than being randomly killed on the street, while walking or driving during your normal daily routine.
Unfortunately, there are very few, if any, countries remaining where human rights abuses are not frequent, so mentioning them in unrelated technology discussions has become non-relevant.
I have a dishwasher so I do minimal washing up. I don't have a dryer as I air-dry my laundry but I could get one if I wanted to. I have a Roborock although I don't use it much since having a toddler.
But folding laundry is the one chore which still can't be automated. Even cooking itself has become much easier with air fryers etc.
No, not really. The macaroni are formed by squeezing the dough through the openings, while the cartridge cases are made by stretching small metallic disks into cylindrical shape (and then stamping the bottom to make it flat and have a proper rim), so it's impossible to repurpose the pasta-forming machines into the casing-making ones.
But it is an urban legend of sorts, along with "baby strollers with chain tracks instead of wheels", and shows how little most of us know about industrial manufacturing.
Please keep in mind that "uncut footage of ..." is the same as those instagram videos where people throw a basketball backwards 20 meters while drinking beer, straight into the goal.
It doesn't prove this is average performance. In fact, if it was, that's not what they'd be showing.
From my skim of their paper, they have gotten robodojo 14% versus previous sota 9%. So its an improvement, but doesnt seem like a reliable model yet, which is fine. I also notice the model becomes worse with more data on the printer refilling task, so thats either a bad run or its a data point against the robustness. The transfer learning results are meaningful. 75% on new complex tasks from <10 hours of demo is really good. I believe this is a 2x improvement over the SOTA. 10 trials per task is standard for these papers but is pretty weak statistically. Memory tasks its weaker at but they have acknowledged it, and I cant see why this arch would have problems w that in the future.
The video cuts to a new shot every couple of second instead of one continuous shot like all robotics research. This makes it hard to trust it. The robot probably failed a million team and they stitched the video from 100 experiments.
And it shows the robot getting into something of a loop when trying to fold the white shorts. I wonder if someone pushed a button to get it out of that.
I'm not sure why some comments here are so pessimistic. I was grinning ear to ear while seeing the video. You're telling me we've finally got robots that can do our laundry? And the model is at least as free as beer?
To quote the lyrics/meme, we used to pray for times like this. This is what robotics should be used for.
There tends to be a lot of low-key (and some not-so-low-key) anti-Chinese xenophobia on HN, so that's likely a big part of it too.
More like pro-US exceptionalism, and any announcement for products created outside of it (whether it's EU or China) is met with doubts and skepticism.
Unless something is from Japan, then it's perfect and any criticism is because you don't appreciate its unique cultural heritage.
There has not been a pro-US exceptionalism tone in any HN thread in awhile. Go to any thread recently and it's pretty anti-capitalism/anti-US.
In all open weight model discussions there is at least some anti-Chinese sentiment
They are likely Effective Altruists highly attuned to AGI (With current LLM's) risks, fearing an LLM uprising and human extinction, which they regard as the world's most critical challenge.
Ironic, given the effectiveness of the open weights models that are distributed altruistically.
the name stopped being relevant a long time ago when they went from rating charities to coming up with horror stories about ai. "ineffective egotism" would be better for what they are now.
I've definitely noticed even more than that - any submissions painting the current USA in a good light are instantly and mercilessly buried (you can check my submissions for examples), while anything even remotely on the "USA bad" side gets to the top. It's really a sad state of affairs.
You may take this as “USA bad”, comment, but let’s take the U.S. life expectancy reaches record high (in 2024) link you posted.
That sounds like a good thing and it is. But it’s not a US story. In fact, other wealthy countries didn’t just already have a higher baseline life expectancy pre-COVID, they rebounded to it post-COVID sooner than the U.S. was able to its already lower level.
So what this actually is, is a story of “global modern medicine good”, and unfortunately, even within that, the U.S. is among the worst performing.
I tried to read the Washington Post op-ed piece you shared on reasons to be optimistic about America’s future. As soon as I scrolled down, the content was replaced by a patriotic fireworks animation and a paywall. Seemed a bit on the nose.
aka this very thread - top comment and most of the entire comment section has been taken over by literally "how dare anyone say anything even remotely negative about China, what's wrong with you".
Is that so shocking on dominantly US forum? I wouldn't expect different situation, people are not 100% objective even if they try hard.
That being said, >95% of the world lives outside US and in recent times (culminating with orange man's second term but started way before) US has alienated literally every single other country, democratic or not. This is what happens when reins are firmly held by unstable people full of pettiness and mental issues, plus visibly corrupt too.
So division deepens as seen also here, and as much as 20th century was a century of US, 21st century is and will be one of China. Every move US makes now just pushes whole situation further into that direction. So folks cheer for their own camp, not some larger common group of aligned friends.
It's strange because my impression has been pro-Chinese comments on HN. I suppose it depends on the topic, the time of day, and how many bots each government is employing on a given day for the key words and their compute budget. This submission seams rather balanced to me.
Is there evidence of any state-sponsored bots at all? It seems like this has become an accepted fact but I don’t see any data indicating it’s actually real.
Pretty sure there isn't.
HN has lots of hellbanned bots lately (I have showdead on [1]) but I never notice ones pushing a political slant. But I suppose nation states conducting a campaign on Hacker News would be much too cunning to make the bots obvious
[1] because many posts get wrongly flagged, apparently in part by trigger-happy users who will flag anything with incidental similarities to AI prose (everyone, PLEASE STOP DOING THAT), in part because HN's automoderation is severe for new users, and in part because plenty of people get hellbanned either unfairly or although they do still write plenty of worthwhile posts. I even rescued someone who it turned out had been hellbanned by a mod by accident.
This is one of the most pro-china forums out there, what are you talking about
Well. I already decided some time ago that I don't want to have my money go into US corporations as long as the orange King is living out his genocidal fantasies (and even without that, I am already annoyed to no ends about Apple, Microsoft, Facebook and Google abusing everyone else so blatantly and violently, including buying legislation - look at the mandatory age sniffing movement before you can access information on the www any longer; this is fascism, and children are abused as fake reason for change here). But I am not all that happy about China becoming so dominant in the whole product chain either. I think the production should be more balanced world-wide, globally. And also, very much so, locally. I am also aware that this is very hard to do, for various reasons, but still if I have a little influence then I really would rather not empower the USA or China any more than they already abuse everyone else here. And both do, by the way.
Genuinely curious, who is Donald Trump genociding? Or am I misunderstanding what you mean by "genocidal fantasies"?
A fantasy is something you'd like to do but don't.
Sure, but the original comment said "living out" those fantasies.
Right, but the original commenter said "living out his genocidal fantasies", which means he is actively performing genocide on someone and it also implies this is something being done with active, excitement.
I'm simply asking who are these people being genocided?
Palestinians.
Palestinians? Israelis are loathe to admit it but that country wouldn't last a year without support- from non Jews!
western elites are currently at odds with the chinese elites and every impressionable idiot is being inundates with antichinese propaganda.
Lets be frank, every single issue we have with the CCP is either also being done by the west or in the works.
chinese oppression of the uighurs is nothing compared to whats going on in Gaza.
china forces prisoners to peel garlic while american prisoners are paid pennies a day to fight wildfires while being prohibited from being paid for those very same skills as firefighters when rehabilitated.
america is quickly workin on legislation to to make criticism of certain foreign governments illegal with very real reprucussons (revoking passports).
At least china has a working power grid and high speed train network.
It's possible to intelligently discuss the merits of Chinese society and also acknowledge that it's a brutal autocracy.
Making some kind of equivalence of the two in terms of the rights of the individual is so ridiculous that nobody is going to take anything you say after that seriously, you do your arguments a disservice by doing it.
The rights of which individuals? You’re familiar with nspm-7, right?
That you can freely talk about nspm-7 on an American forum without fear of being disappeared shows that America and China are not at the same level.
Yet.
"Yet" is the keyword. it won't be long before we arent' any different from china at the rate we are going at and I think a lot of us would be better off preventing the west from turning into an oppressive autocracy instead of judging china.
Saying that the US and China are currently the same is admitting defeat and giving up the fight before the fight has actually been lost.
The fact that this text still exists alone is enough to prove my point.
Nation wide NIH.
It’s not just HN. The USG has been manufacturing consent for war with China (trade or otherwise) in the US population for decades. Same goes for Iran.
Every single thing that’s come out of China in the entire AI/Robotics space lately has been a massive net positive for the world
Open Source is a good thing, regardless of where it comes from
It's also relevant that China isn't open sourcing results because that's something China will do forever on some philosophical grounds. They do it because it is currently in the best interests of the Chinese state (and the world, as it turns out, coincidentally) but as soon as that stops being true there is zero chance it'll continue to be open source.
Yeah the US needs to step up with its Open AI research here, they should start some kind of "Open AI" collective to release models free to the world, do it all open like
So I'm supposed to hate this because of what China MIGHT do in a future neither you nor I can predict?
Is there someone saying you should hate something?
If we want open source technology we should recognize that the current arrangement is non sustainable - the moment china doesn't have an american AI industry foil to play against it'll immediately close. It's up to the US to build more open source models. We already own the content used to train them, collectively. We should own the results as well. Make the big AI companies release weights.
Reminds me of the american reprint market for books back in the colonial war area.
Eh, I take what I can get. A lawyer doesn't argue that his client is innocent (though that might be tha case), because of some innate sense of justice, but because he gets to paid so.
Yes, this is the correct attitude. Fire is useful, but don't forget it burns you when you stop paying attention.
I keeping saying this, the best AI / LLM coming out isn't another question / answer chatbot or automate my email and essay.
It is Robotics! Finally doing my dishes and laundry for pete sake! And hopefully help me to find stuff I keep forgetting where I put it.
I have so many airtags because otherwise I'd lose all my shit.
This super interesting link showing things that were impossible even a couple of years back, sitting below the day old news about Kagi's me too browser tells everything about HN audiences. West will fall behind in robotics and mostly because of the wrong priorities of its VCs and developers.
But can it goldplate oysters?
The singularity is literally here right now! Some people will never be happy! This is easily the most amazing thing I have seen in my life! These things will be taking care of and raising babies very soon!
> You're telling me we've finally got robots that can do our laundry?
No, we don't.
I wish we did, that would be amazing.
There have been demos like this over a decade. But they are just that, demos. Demos done in a controlled environment.
If there was a robot I could buy that could effectively fold laundry or do dishes, I would gladly drop 50-100k on it, but such a thing doesn't exist and likely won't for a long time.
You would honestly pay 50-100k on a robot that just folds laundry or does your dishes? In the US you can get a maid that cleans once a week for $150-250 and would do a much better job. By the time that cost would equal this robot, the robot would be obsolete, not to mention any maintenance costs.
Also dishes are the easiest chore to do around the house, especially if you clean them after use instead of letting them pile up.
I honestly think I would. I have a maid come every two weeks which is of course not often enough for dishes/laundry. A maid that only came weekly couldn't stay on top of dishes either but they could do laundry fine sure. I do a lot of cooking and I find constantly doing dishes and loading/unloading the dishwasher to be a huge never-ending pain. A device that costs 100k and would free of me that sisyphean task forever would be a godsend. Also you can have the robot work anytime whereas a maid is probably on a schedule.
And not to move the goal posts, because my original post did only mention laundry and dishes, but I assume once a robot reached a point where it could do laundry/dishes effectively without supervision, that it could do a variety of other tedious household tasks. 50-100k would be well worth it IMO.
The pessimism in the comments here is entirely the reflection of the times we're living in.
Technology got weaponized against the average citizen.
It has become a synonym for unemployment, enshittification, mass surveillance and social erosion.
If this video appeared in 60's or 70's, people would be in awe imagining the future where robots do everything.
"the future where robots do everything" and "unemployment" are the same thing
Yes but unemployment where a small cadre of rich elites own the robots and the rest of us are serfs is very different from the public owning and benefiting from the robots in the form of strong social services and UBI.
The Jetsons vs. Elysium.
If everyone is a serf who buys the products that the elites want to sell?
This seems more likely than racism, or us exceptionalism as mentioned above.
People are and should be skeptical of this. The natural response would be
a) It will put me out of a job
b) I won't be able to afford it anyway.
Those of us who liked the open web and P2P and hacking on projects all the time are seeing these models as a way to rebuild the internet in a shape that pleases us. To build anything and everything we want. Sans opportunity cost.
And that's the real value - you don't have to burn the candle of your life away to get what you want. You'll have so much more time now that you aren't AST wrangling.
Opportunity cost is vanishing (or at least the cost is multiple orders of magnitude smaller). And that's the most exciting thing that's happened in tech since the internet.
I'll probably be able to make my own phone operating system.
I'll probably be able to decompile N64 games, port them to Rust, combine them, extend them. In a weekend.
I'll probably be able to make my own music management platform while simultaneously working on a few robotics projects.
A find my phone button.
Automate everything.
And that's maybe a month of projects.
The whole thing feels like that the US collectively isn't feeling confident about itself in the coming creativity-limited world that AI/AGIs supposedly bring, and that it's not happy about that pessimistic rational projections. I don't know exactly why wouldn't it be or why nothing seem to be done about it, though. Both look to be fundamentally solvable problems to me.
> You'll have so much more time now that you aren't AST wrangling.
So much time because you were fired and have nothing to do. Nothing is as good as oodles of quality time worrying about funding your family's needs. It's amazing.
Robots will take jobs from massive amount of low income people.
AI will take jobs from a massive amount of avg and high income people.
The disruption ahead is getting pillowed right now due to physical limitations of progress like amount of energy, building DC, doing research, training models (its still takes 1-3 month).
Also a massive buffer will be the boomer going into retirement.
Depending on the speed of all of this, we might get a lot poorer before we get abundance everyerwer
Yep, this is a good read and feels like the correct take.
> Depending on the speed of all of this, we might get a lot poorer before we get abundance everyerwer
not "might", will
and it's a hell of a leap to assume massive abundance will follow -- for someone, perhaps, but not the average person
At some point it will be cheaper to give everyone bread instead of paying for your private military.
But yeah lets see what going to happen.
But there's no feeling of power with giving everyone bread. Billionaires seem to have more than just an addiction to big numbers, maybe seem to want to rule over others.
You have no evidence of this.
Also: you can literally make incredible software that would take half a year right now for $200/month. Or free if you run local open weights.
You can do branding, marketing, design. You can create comics and movies and video games.
You can make an AI tool right now and get customers today. You have no idea how wildly popular this segment is. How much people willingly engage with this stuff and want it.
Yet you choose to disengage and complain about a future that hasn't happened.
It's more likely that firms will need to hire more to stay competitive. Or that workers will fire their old jobs and break out on their own.
Something I used to think was impossible: competing with Google, Apple, Stripe. Not anymore. I think you can compete with anyone.
I saw robots doing this in tech demos 15 years ago. And even those robots weren't doing anything new. Like all AI models, this is just AI repackaging and rebranding of things that robots have already been programmed to do using more classical control methods. The idea to do everything via AI was of course thought of for a long time, but the conclusion has been we have faster and more reliable methods to do the same thing with less compute, and that's still true.
During my undergrad I worked on a project where we set off a bunch of walking robots, recorded their walks, and then used the data as the basis for a learned walking controller. I see this Xiaomi model as pretty much like that.
It's a nice technology and maybe could lead to better folding one day but unfortunately it's not "finally we have robots that can do our laundry" (yet).
Yeah exactly. I've been seeing demos like this for ages. I'm kinda shocked at the lack of skepticism in the comments here. The hackernews comment sections have changed a lot these past couple years...
> You're telling me we've finally got robots that can do our laundry?
I’m still waiting for it to peel, slice n dice onions so that I do not have to cry while doing so, and not pay $8k for that robot.
Did you see any of the MSE numbers? around 50% success rates in the tasks.
Anyone in robotics can tell you that building an 80% solution is doable, very fast. Covering the last 20% can take your entire lifespan
> You're telling me we've finally got robots that can do our laundry?
Did you watch it actually fold that shirt? Super slow and bad at it.
Is a Chinese robot vacuuming and mopping my floor every day? Yes it is. Could I do it better manually? Absolutely. I still prefer the robot to do an ok job than to do it myself.
The difference is that vacuuming seven times as often can compensate for a less effective process. Folding 7 times doesn't eliminate wrinkles.
No one cares about wrinkles anymore. Come on let's be real.
Apparently not given how many comments in opposition I got.
Most people in software misunderstand how physical goods work. When making machines you don't care about getting everything perfect all the time, you care about staying within tolerable limits all the time.
How bad do the folds in your laundry need to be for them to be no different then just being dumped in a pile? As long as we are above that point the robot is useful.
I think there is a stark contrast between the plus or minus micron level tolerances, and whatever is being churned out of LLM ('not getting it perfect all the time')
I don’t need it to do it faster than me, I just need it to do it instead of me.
Plus it’s obvious it will get better.
> Super slow and bad at it
Eh, this still has value for me if it can do it reliably. Leaving a robot to badly fold my t-shirts overnight and then badly put them in my closet the next day is fine. I don't need my casual wear beautifully folded. But I do need it organised, and that's apparently a chore I'll just never learn to like doing.
Absolutely!
Also if this scale anywhere close to LLMs in 2 years they will fold at super human speed and much better than me.
The real question is whether a hardware platform today would be good enough.
The model is software on a server or in the cloud, so the real question is are there any show stoppers from the hardware for going faster?
Physics. Right now the clothes folding is computationally limited, but assuming we get faster computers to get them folding faster, gravity and air resistance are still going to take their time.
Exactly. I will pay money for a robot to take laundry out of the washing machine and hang it to dry and then fold it and put it away or just sort it by size. This is a chore that my wife and I have to do several times every week.
My solution: stuff the clothes from the washer into the dryer, then stuff them into my wardrobe. I buy clothes that don't get wrinkled much even if simply shoved in the wardrobe and if they do, I tolerate it and wear them anyway.
Ha! You remind me of something I used to tell my colleagues back when I held a regular job. They would compliment me on my highly patterned shirts (eg. A very small floral pattern repeated). I would explain that a shirt like that just requires low spin in the washer, and then drying on a hanger. Whatever wrinkles are left are almost invisible due to the pattern. With collars and cuffs I’d just literally manually stretch/tension them while they were still wet. Bada bing bada boom.
Sorry, but if you're buying clothes that have manual drying steps, it's gonna be a hard sale to the general public.
This is a mostly American perspective that I find endlessly amusing. I only have a handful of clothing items that say that tumble drying is fine on the tag, and most of my clothing is very casual wear (t-shirts, jeans, sports clothing). Dryers eat up clothing like it's nobody's business. Plus, electricity costs money while the sun is free. And while yes, the time to hang laundry has value, it takes 10 minutes to hang a load. If it rains, I hang the clothes indoors with a dehumidifier which is cheaper to run than a dryer and dries clothes in a matter of a couple of hours.
To be fair, I have quite a few of these, and they inevitable just get shuffled over to the dry cleaners.
I was referring to shirts specifically. If you put any kind of cotton shirt in a dryer it’s almost certainly going to need ironing. Waiting for a shirt to dry naturally shouldn’t be a big deal.
So what? it can take hours to do it and I can focus on something else. If these robots actually become available on affordable price I would buy one without blinking. It would totally free up my weekends, no more hours wasted doing laundry, cleaning the apartment etc.
My dishwasher is also super slow and bad at it. But you know what? I still prefer that over doing it myself.
Your dishwasher shouldn't be 'bad' at cleaning.
Here are some tips:
https://youtu.be/jHP942Livy0?is=B_CGi4shCuiwQLvR
It dulls sharp knives, and is very agressive to other materials too.
It's kind of crazy what goes on in that machine. It basically turns into a mini-Venus in there.
The very first car was absolutely awful. Super slow and bad at it.
Slower than a horse even, why even bother
Xiaomi's first car, on the other hand https://youtu.be/KGW8bQtcpJg
I am worse than the robot and I don’t mind if it takes 3 hours to do it.
It can do it over 8 hours while you're sleeping though.
and cost more than a maid
Sure but that holds true for most new tech. Expensive and slow but it can only get better or stay the same. I'm thinking it's only going to get better.
You're missing the forest for the tree.
I can also vacuum my floor 3x faster than a robot vacuum. But so what, I'm not there when it happens.
"Did you see the output of that Chat GPT 2? It's super lame code"
The outlook on this, as so often, depends on personal finances.
> we've finally got robots that can do our laundry?
Until it can check all pockets for tissue paper...
My wife already doesn't do that
> I'm not sure why some comments here are so pessimistic.
Actually, it seems like 99% of the comments are glazing this with zero skepticism actually. If anything the comments are too enthusiastic and lack the measured discernment that was more common on HN 5+ years ago.
> I'm not sure why some comments here are so pessimistic.
Time and time and time and time and time again, technologists are always naïve to never realize that the road to hell is paved with good intentions.
I am expecting the opposite effect to happen with the actual use of robotics. Mass displacement of every sector of labor.
This is not ruled out.
Mass displacement of every sector of labour. Dissent policed by Boston Dynamic's robodogs. You will own nothing and be happy.
> This is what robotics should be used for.
I disagree. We should use robotics to extend our reach into hazardous environments, enhance our physical capabilities, or to provide a steady handling of things where it's absolutely needed - doing laundry in my household is none of these.
Folding laundry takes me 10 mins every 10 days. I organize my closet once in a while according to my sense of fashion, and now I'll just need to keep that structure. I don't always like to do these chores, but I have no issue in acknowledging them as part of my responsibilities. Why would I even automate[0] that?
Robotics is costly, and I see no way how that would change anytime soon. That isn't a flaw of some kind either, it's rather just one of the limitations that needs to be accounted for, neglecting that would be detrimental to the future I envision.
Don't get me wrong, I love robotics with a passion, as engineer I like to tinker, and I'm a big admirer of FESTOs bionic research[1]. It's just that the field suffers from an "overpromised and underdelivered" aura as far as I can remember, and I remember the presentation of ASIMO[2] 26 years ago.
Some people would like to see the field hyped up, focused on the consumer like every other field in tech, probably for financial gain, but I personally see the value in its downplay. To be honest, I love these reality checks, as I see them as a valuable part of the learning process all together - a journey for what it means to automate, and what it means to define as human natural chore.
[0]: https://xkcd.com/1205/
[1]: https://www.festo.com/us/en/e/about-festo/research-and-devel...
[2]: https://en.wikipedia.org/wiki/ASIMO
Sure but the robots that extend our capabilities often start by replicating our current capabilities, even if those are as simple as folding clothes. It provides a baseline for the next generation of robots that can do more.
You could replace "robotics" with "AI" in what you wrote above, and it would sound like what people were saying about llms when they first came out. Not trying to glaze one way or another but historically this take is outdated.
A subset of the people on HN hate AI and yell "slop" at everything they see. It's even worse if you go to other social websites.
Just ignore these people. If they don't use AI, they likely won't remain practicing in our field for much longer. They will naturally self-select themselves out of relevancy.
I was Claude coding and got hit by the 3 AM reset bug. I went on reddit and saw hundreds of other engineers hit by the same. We are different people than the AI haters, and there has never been something so powerful it's kept me up past 3 AM every single night working beyond every fiber of my body telling me to sleep. I feel empowered as if I command an army. And there are millions of us. The sheer might of all of this will win out over negativity.
I've built multiple successful businesses in the last year. If you're doing this with any sort of interest in self-employment you don't need your employer anymore.
I have never been happier. This is the best future, and I feel lucky to be living it. It's what I dreamed about when I was a kid, and it feels magical. As far as I can tell, the future has no limits.
That last paragraph is particularly worrisome. I worry about other dilusions you have imbued.
This is not meant as an attack or to sound cruel but you are suffering from some degree of psychosis. Please look after yourself.
It's always funny seeing people that write like this as if they are somehow exempt from the future we're seemingly rushing into. Bro, you prooompting yourself into unconsciousness is not going to make you not obsolete. Unless you're being sarcastic in which case poe's law.
Yes I think perhaps he's trolling and I fell for it
Working past your personal bedtime, part of an imaginary movement of millions might, ignoring the negative aspects of AI, empowered like commanding an army. This frankly sound psychotic, and it's exactly why uncontrolled AI usage is seen so critical by many.
And all other devs will leave the field just because they prefer to work differently? That is not only not realistic, it is violence fuel.
You will have to work long nights when you try to build a business of your own while continuing to work a regular job.
Talking about psychosis and "violence fuel" while handwaving at negative aspects of AI does not seem substantial.
I would be more curious to hear details about how OP built the claimed businesses.
I will not see this info here if people who build things and are enthusiastic about AI are driven away by downvotes and negative comments.
The frustrating part is that these posts will never post any details about the business, despite the fact that it would immediately invalidate most criticism.
It seems to be a similar phenomenon to productivity "gurus" who write and evangelize polyphasic sleep and "grindset" and so on - when you ask what they're using these superpowers to actually accomplish, they're very quiet. Or it's "coming soon." Or "I'm not going to waste time explaining to someone who obviously doesn't ~believe~." It quickly becomes religious.
You're not going to be able to keep up with the market if you're not using AI.
100kloc in a day is table stakes now. And that's just where we are today - imagine a year from now.
You can't continue to be a horse cart driver and get paid for it. Or an abacus user. We're moving past it now.
It is not a sign of productivity to create LOCs. Especially not 100k.
Even if you're Fabrice Bellard, you can't compete with how fast these models are. They'll run circles around any human.
I don't know what else to tell you other than you should try the models again and update your assumptions.
A year ago I thought LLMs were trash. This year has been a completely eye opening experience.
Do not be so stubborn that your opinions are unshakable. Something exceptional is happening and it deserves just a little bit more of your attention.
it's fine it can run in circles all it wants, i just want a calm, stable life with purpose (no running an llm ain't it)
Work as a nurse or teacher then, or some cushy government job.
One year ago LLMs weren't trash, in one year you'll find that they aren't miraculous either
> there has never been something so powerful it's kept me up past 3 AM every single night working beyond every fiber of my body telling me to sleep. I feel empowered as if I command an army. And there are millions of us
Yelling "slop" at a clothes-folding robot is stupid, but reading this kind of shit makes me want to embody Ted Kaczynszky and go live in the woods.
You forgot the murderous psychopath part...
Ahah you AI bros can't stop talking about Claude and llm, and telling stories of how great it was that one time, even when they are not the topic at hand.
> I've built multiple successful businesses in the last year. If you're doing this right you don't need your employer anymore.
Right
Rolled my eyes at "I've built multiple successful businesses in the last year". This person is clearly in some sort of manic state.
We nearly destroyed our planet to get there. Technocrats here applaude the new technological achievements that make life a little easier (less boring?) for the top one percent while willfully ignoring the associated costs for everyone else. Manufacturing these at scale will only make things worse. We are approaching peak decadence ever faster.
The top 1% already have humans to do things for them though.
Wall-E tried to warn us
I am of course impressed by the model and the robot. Great stuff,
But the downside is that won't we all just turn into the obese, stupid humans at the start of the WALL-E film?
Mindfulness teaches us that we can enjoy the washing up, the laundry, planting seeds, tending the garden. It connects us to something. Without this will we end up with a rise in mental health issues, because for one thing people will just end up on social media in the time they save.
Just a thought.
Ironically, the people who can afford to have in-house help to handle things like cooking and folding clothes - the rich - also tend to be fitter and more active
Maybe if you can shave an hour every day from chores, you will actually have the energy to enjoy physical activities
Probably not, as the washing machine, dishwasher, car, computer, etc didn't cause that either.
edit for nuance: Sure, it did for some people I suppose. But they are the exception.
I often wonder if they just made us do more washing, use more cups, travel further in the same time and get more expectations from our boss though.
I've got 2 kids. I'm always doing laundry and loading the dishwasher. How much of that is just higher expectations of cleanliness? Back when washing clothes involved going to the river all day, kids would get sewn into their clothes for winter. These days they have 5 outfit changes a day. I'm certain that my dishwasher means everyone just grabs a clean mug/glass/bowl rather than just rinsing a dirty one. I use a computer all day to get stuff done. But that just means my boss expects more stuff to be done. The baseline shifted.
A few hundred years ago people mostly lived near where they were born and didn't travel much. Now we get in our cars to drive a few hours to visit our family and friends. But they only moved far away because the car enabled it.
Obviously there are benefits - especially for those who find dirt upsetting - but have we actually changed the balance of leisure time very much?
I mean I like the much lower rates of dysentery and untreatable skin and dental conditions...
People have an extremely rosey view of the past shaped by the fact that period dramas are acted by modern Hollywood actors.
Or we can use the free time to go to the gym.
Anything that frees us from labour is good.
Unemployment frees us from labor too but it's not exactly good either, I'd rather avoid it.
> Anything that frees us from labour is good.
Including labour itself? All we're doing is optimising the amount of time we can dedicate to work.
This discourse reminds me of the handful of Christians who thought we needed Corn Flakes and uncomfortable clothing to stay chaste and that electricity made us stray from God.
Chill. Your brain is the product of the past. Reread that: your brain is anachronistic and outmoded.
We get to live in the future now. Count yourself lucky. You're one of the few that got to witness the birth of flight culminating in landing on the moon, except in this case your floppy disk became your personal shopper and stay-at-home Michelin star chef and part time driver.
It's okay to feel a bit of discomfort. It's hard to believe this is actually happening in our lifetimes, but it is! And that's so cool.
So sit back and enjoy the future.
For a bit of context there are some classically very hard things going on in this video, so if you're not impressed, be glad you didn't have to try to build this 10 years ago.
- Coordinating two hands
- Mobile body
- Deformable objects
- Thin affordances (bag zip in particular)
- Multi-object single grasp
Really a host of things that had individually been tackled as either perception or manipulation problems and probably led to more than one person's PhD here on HN.
Watching the two-handed robot has me thinking that a third hand, perhaps with a different set of capabilities (like a tentacle grip) might make many tasks easier. I realise that more limbs = more problems, but even so, having additional ways to hold and turn things would be a big help.
My thoughts exactly, but to train it you'd need a lot of data with someone manually operating it. With just two hands it's enough to have a lot of footage of humans doing it. Maybe down the line the robots will be able to self-improve or someone will come up with ways to optimize this..
SciFi novel "The Mote in God's Eye." has aliens with three hands.
One "gripping hand" kinda coming out the gut, and two dextrous hands for detail work. Led to abuse of English language with: On the one hand... On the other hand... On the gripping hand...
Which is reall just a patch for the problem of "On one hand… On the other hand… …On the OTHER hand…".
It's worth noting that the one-other-gripping pattern seems meant to convey two considerations and end with the strongest of three.
The English one-other pattern doesn't necessarily require the second consideration to dominate. They could be equal, or they might merely introduce concerns that haven't been evaluated to conclusion yet.
Alas, one-other-gripping doesn't seem to generalise to more than three concerns.
Anybody who worked on a car knows an extra hand would always be helpful.
I feel like there'll soon be a market for AI controlled enhancements like limbs. Neural interfaces seem to be a long shot, at least for now, so some kind of AI that senses the environment and intelligently react to it seems like the next best thing
I'd settle for something to just hand me required wrenches/bits/heads/etc while I'm under the car. This actually shouldn't require much.
tentacle grip.... giggidy giggigdy gooooo
We need Amdahl's law for limbs
Yeah, folding sheets requires 4 hands. With a large distance between each pair.
Or just two to four robots working as a team.
Coining the term: Slopfold
When you Robot folds your clothes but it's kinda wrinkly and sloppy but you accept it anyway because it's easier than precisely folded and pressed clothes.
I have been slopfolding since before it was cool
I've been slopfolding since back when it was just called "dumping it in the wardrobe".
Like outsourcing my thinking to the internet, I too did these things before it was cool.
Isn't that the story of <any> technology? When the coffee machines first came to be, Barista was definitely a better option. However, through the evolution, in many cases today machine does better job than barista.
Better analogy would be that even though the coffee machine is still worse, we choose it because it's more convenient (at least I think that's the case).
It's cheaper. That's, like, the only reason ever the mass market adopts any alternative.
And usually better, too.
Anecdotal, but the best machine will never be able to do something as good as the best human in most circumstances. That is because, unlike machines, humans have passions and character that subtly imbues itself in the end result. This, of course, isn't limited to coffee, but to anything a machine could do that isn't strictly tied to exercises in logic (for example, computers will always be better than humans at most games because they can be analysed logically and acted upon)
> for example, computers will always be better than humans at most games because they can be analysed logically and acted upon
This made me wonder how we can imbue some of the the reasoning in your post to it. A game that values human ingenuity
I feel like, despite not being my (and presumably most of HNs, given the general demographics) cup of tea, sports are largely that. The players are skilled but also uniquely themselves. The world cup wouldn't be nearly as interesting if it were just unthinking robots fighting to see who can run around and kick things the best
My drip coffee maker makes me coffee that I find more enjoyable than stuff from a barista at a fraction of the price.
saw the way it folded the clothes and thought "better than me" :/
I'm not sure what the :/ implies, but if you're disappointed by your own folding techniques, spend a bit of time on it, look up some guides, and fold clothes with purpose. or something.
(Rise up against the machines! Folding your own clothes is an act of resistance!)
I use it for "disconcerted", but I guess any situation that would result in a diagonal mouth irl would be fine.
I've always taken :/ to be a kind of hopeless dejection mixed with apathy. In this case, it seems to convey their disappointment in their skills as well as the futility of improving
I read it as an acknowledgement that while improvement is possible, the requisite effort will in reality never be expended
Folding clothes is already a compromise between leaving them unfolded and hanging them
Funny how hanging them is also a compromise between leaving them unfolded and folding them.
I suspect you might be jumbling up your smalls with your origami.
How privileged was your upbringing?
That comment could easily have been made by an orphan from the poorest part of the US who signed up for military service.
I don't know why you would assume privilege from a comment like that. If anything, it's the unprivileged kids who would know that you'd get your hand slapped for folding clothes like that.
Speaking as a relatively unprivileged kid from Southeast Asia, my parents were so busy working they just didn't have time and energy to care about me, let alone how well i fold my laundry.
But that's just one data point.
When AI is a the forefront of scientific breakthroughs and all human endeavor is painfully minuscule compared to AI, we'll still be calling it slop as we all hypocritically generate more and more slop.
God this whole comment section is weird.
It's like all the people who are so proud they don't use dishwashers.
Nobody is proud of this. Where the hell did you get that idea?
I think weak and scared people tend to not want to face the truth about AI. I said the above not because I support it (I don't), I said it because it's the god awful truth. All trendlines point to steadily improving AI, and tons and tons of people continue to call it slop even though it is clearly becoming better.
My question for you is, in order to sleep at night do you have to lie to yourself? Do you make up stories about how AI will never be as good as human intelligence and do you assume anyone who says the contrary is an ardent supporter of AI? Have you thought that... maybe nobody really supports AI, they're just telling the truth?
Weird my ass, just stupid people who clearly assume the wrong thing.
No definitely weird.
It's happening so fast now.
AI will absolutely dominate our future. I believe we have only scratched the surface of what is possible.
Maybe I'm pessimistic but I always thought Bill Joy was prescient.
https://www.wired.com/2000/04/joy-2/
this is what excites me most about the probable AI future. househelp is cheaply and easily accessible for most people in the east, however being able to afford a cleaner or maid in the west is a luxury. i spent a ginormous amounts of time every week keeping my house in shape. i look forward to the day when i will be able to carve out that time and get it back for myself.
Househelp isn't cheaply and easily accessible for most people in the east, a large portion of people in the east is just poor but you don't consider them in 'most people'.
I think anything more advanced than a Roomba won't be cheaper than a maid for actually cleaning your house.
I wonder, what will you do with that free time you will crave?
Work more for my boss and landlord so the fiefdom can survive.
And pay for the maintenance costs of my new robotic servants! At least now I can finally feel like my own masters, ruling over these robots. /s
i want to read more for the most part. my reading list is evergrowing.
shameless plug: https://whereistejas.com/inbox/
(PS: people who comment about the faults in my website will be downvoted. im sharing my reading list, not looking for a codereview)
Here's 10 things just starting with "A".
Amateur astronomy
Artistic painting
Audiobook binging
Aerial yoga
Archery practice
Aquascaping
Antique hunting
Acoustic guitaring
Alpine hiking
Astrophotography
.. Need more?
The uncut footage is so awesome.
Of course, you can mention all the things they can't do (what if there are stairs, what if the doors are smaller than the robot, etc) but it's mindblowing to see what they can do - and also the chat panel is really nice and makes it more interactive.
Designing a non humanoid robot is the way to go. The design of new tools and appliances for hybrid human/robot use will eventually follow: the zip line for the shoe bag required way too much precision work.
I think this perspective is now demonstrably false. It was basically the defacto position of most roboticists until a few months back. But we've already seen the cost of humanoid robots fall precipitously (even without the kind of mass production that genuinely useful home and real world intelligence would bring).
Which seems more likely, that cheap humanoids will continue to improve in dexterity, battery life, strength etc. Leveraging the vast investment in research into real world task training. Or - that we'll redesign millions of everyday objects to fit the limitations of current task specific robots?
If you answer the latter, I've got a whole bunch of mid century quirky labour saving inventions in the back room I'd like to sell you.
It’s not demonstrated yet as we still have to see a robot (humanoid or not) doing mundane tasks in full autonomy, efficiently and convincingly.
It still looks inefficient to have a humanoid robot vacuum a house by using an ergonomic tool designed around the human body.
I can't believe someone shares my point of view. We have already many non humanoid robots to do boring stuff (underwater maintenance, stuff like that). For some reasons people tend to think that for robots to do stuff close to humans, we need human like robots.
We do not. I want a swarm of insects-like robots to fight mosquitos. A snake one that fixes my pipe. A spider-like robot that cleans surfaces. Etc.
the cube shaped one that I put my dishes in.
I will suggest mix of both: some everyday tools are already built for being used with hands and legs (e.g opening a door). But indeed adding some different grip types, additional arms, etc, could be helpful.
Exactly what I came here to say. A robot doesn't have to look like a human. Just a few arms that does the job neatly is more than enough.
This is amazing! I can see in a few years having Home Assistant use an LLM to coordinate my home while I'm away. Tell the LLM to make an inventory of my home and then send the vacuum and the robot around to clean everything up. A multimodal LLM can inspect pictures for items that need to go in the dishwasher, the laundry or needs polishing and reorganise my closets to best allocate space, make the beds, take care of the plants. It has all day to do it while the solar panels provide ample energy anyway.
As a house spouse I want this! I _hate_ folding laundry and always procrastinate on it. Doesn't even need to be able to load the machine like they additionally show, "just" folding it is enough for me!
Does this require a given hardware platform, or for example, could I run this model with a cheap mini robot with a single camera (i.e. good enough without depth), tracked treads, and an arm?
When the page says "embodiment-free (UMI)" it means they've used a standardised gripper and camera arrangement [1], while aiming to be agnostic to the type of robot arm that gripper is attached to.
[1] https://umi-gripper.github.io/
I’d imagine the training is hardware specific
I love this future. A robot that does all the chores relentlessly of the house while you spend time just working on the stuff you like.
I think this robot is really well-made! Xiaomi seems to be quite controversial these days, but personally, I still really like the brand.
I have Roborock (Xiaomi) automatic vacuum and it's fantastic technology. iRobot had a 10 year lead and did nothing with it. The robot is competent at avoiding clutter, clean the mops itself, leave my floor sparkling and so far has been extremely reliable.
Got two Xiaomi phones in a row, one in 2018 for 250 euros (note 8 pro) and one in 2023 for 300 (13T) and I liked them a lot.
My gripe with the second one is lowish battery life, it always bites me on longer trips when my gf's Pixel 8 lasts considerably more, but was also way more expensive.
Will check battery reviews better next time I shop for a phone, I don't care much for the rest.
I just can't deal with all the ads on Xiaomi phones. Switched over from Chinese phones, still remember my JiaYu G4 fondly, to Samsung few years ago where the situation is a bit more bearable
I never saw an ad except when opening Xiaomi own apps which I don't use.
Anyone have any recommendations for robot hardware where I could test this open-source model? It would be a cool/fun project to get it working on a little robot.
Always remember this is the Model T from Ford, imagine what they will do in 10 or 100 years from now
Exciting times we live in
This is very cool, and the paradigm that scales to bringing truly useful robots to virtually all domains of life - just as LLMs have scaled to that degree across virtually all domains of knowledge work.
I picture a lot of perfectly tended gardens growing food.
Please help me understand. Is it more effective to have a general model like this instead of a state machine that leverages other models and a deterministic set of code?
feels like the American domestic manufacturing is done, there is no hope to save it.
How are these models compared to world models by Waymo and Tesla, assuming their models are trained for recognising objects that can be further generalized for everyday chores?
Compared how? Architecturally? In training methods? In performance?
The remarkable thing about this modern "full VLA" approach is that it doesn't train for an explicit "world model" at all. There's no factored world state, no pool of trajectory rollouts being sampled from. Instead, everything is rolled into the same VLA stack that does all of it internally.
The world model exists, but it's implicit - the VLA learns one, but only because it helps its training objectives. Trajectories are planned ahead sometimes, also implicitly, for the same reasons.
It's a very "bitter lesson" way of doing things.
As far as I know, Tesla's stack is somewhat similar architecturally - a highly integrated end to end model - but the training process for FSD does include factored "world modeling" side objectives to force the stack to develop driving-relevant internal world representations. So it's in a middle ground between a classic "world model" system and this flavor of VLA - albeit surprisingly close to the latter.
Whether the world modeling side objectives are principled and justified or a leftover from the days of insufficient scale is unclear.
How dare Xiaomi post a realistic picture of where modern consumer robotics is at.
They will never become trillionaires if they don't pretend they are making a million humanoid robots by end of the year.
Really interesting.
From even before the RAM shortage there was noise in the robotics world of sensor and actuator shortages due to them being bought up in bulk by some of the major AI labs.
The first few second of the shirt folding arms trying to put the t-shirt properly on the table to start folding it are at the same time amusing and scaring.
What if the thing retrieves some anger-driven post from Reddit from its training data set and start literally to flip the table??
Neither the Github nor the HF repos have anything on the XiaomiRobotics-1 model or datasets :-( They have a bunch of assets (datasets, code, model, etc.) for XiaomiRobotics-0 however, but I don't know much about that family of models.
> Code: Coming soon
> Model: Coming soon
Obviously, XiaomiRobotics-0 was their previously published model, which must have been weaker.
I assume that "soon" means "a week, or two".
Pretty cool!
Only 10B parameters if I correctly understand some of those histograms, which is tiny in LLM world! It bodes well for the future.
The training videos seem to show human hands using the robot grippers to execute the tasks, which must have been a fun but tedious process.
Yes, it's truly remarkable how small modern VLAs are.
Even Tesla's FSD AI stack is small by modern LLM standards. Once the hardware is there to enable real time 100B inference on endpoint SBCs? We'd see gains in AI robotics from that alone.
How well do models like these generalize to robot embodiments they weren't trained on?
Why cross-entropy follows a curve is fairly trivial to understand:
https://www.youtube.com/watch?v=GlYgs6v2YfU
In general, motion planning multi-axis systems is NP-Hard. Have a great day =3
WAIC has been wild this year. We got Kimi K3 and Qwen 3.8 announced. We got robots boxing and open-sourced robotics foundation models. And we got the announcement of WAICO.
At what point are we gonna start seeing American firms attending WAIC?
I'd like to see a serious HN discussion thread about WAIC first.
In the domain of hardware, what is the equivalent of a harness? How well would this generalise to any robot, versus one that is specifically built the way Xiaomi did?
Considering that they pre-trained on non-robot data, and mid-trained on data from other robot types? I suspect that generalization would be decent for most similar platforms - things with 1 to 2 arms and 2-finger hands.
You'd probably have to fine tune for good performance on a specific arm type. But this model isn't trained against hard assumptions of what the underlying hardware is.
I think this is the most satisfying video I’ve seen of a robot doing laundry. Not sure if it’s the camera angle, the 3× speed, or the music.
I watched it on mute and was still satisfying, so I guess it's not the music.
The uncut movie is at 2X speed.
You can easily save it on your computer, and then play it at 0.5X speed, to see it like in real life.
It was still satisfying.
How are there still people who look at all this advancement and the only thing they can utter is "it's not there yet?" It's clearly here, and if it's not to these nerds' expectations, it'll prob meet it in 4-6 months. This was cool asf.
Because you see a robot do one thing, doesn't mean it can do other things. We've had good robot demos for decades.
For all we know they could've spent 5 years and 10 million dollars on getting this robot to fold laundry. That doesn't mean it'll be able to mop the floor better than a Roomba in a decade.
No we haven't. Back in the 90's Honda's Asimov could barely walk. Now we have robots managing to run marathons. Clothes folding had to be played back at 50x speed because the compute just wasn't there yet. It still is barely there, some of those videos are sped up. They've probably spent way more than 5 years and 10 million dollars developing that robot. Roomba already came out with a mopping robot, why would you want this one to do that?
It's probably because robots have been doing cool asf demos for half a century and they're still mostly just demos. Elon has made a lot of money greasing the 4-6 months away crowd.
Would love to see a load/unload dishwasher benchmark.
Cue the variations of:
"No wireless, less space than a Nomad. Lame"
I hope I can use this sooner
Cool video but it forgot to pack underwear in the bag!
Of all the tasks I do, folding clothes takes a ridiculously small slice of my time. Is this and all the other robotics videos aimed at H&M staff?
They're trying to show their investors a butler robot that can do all types of household chores.
Folding fabric has traditionally been very hard to solve with robots; the ideal places to grab the fabric might be hidden, the fabric patterns are repetitive, and the material flops around, tangles in the grippers, and brings other pieces of laundry with it.
Successfully sorting and folding a basket of laundry shows they've got a control system capable of performing some quite difficult tasks.
They also use a large ring attached to the first zipper to help the robot grab it. This is not a realistic zipper, but I do appreciate them being honest and showing that grabbing small objects is still not perfect
All you need to do to understand what task is more complex (for a robot to solve) is just to ask a 6-8 years old child to do it. It takes days before they "understand".
So it cannot be that simple (for the reasons you mentioned).
As a parent with 3 kids and both my wife and I working, folding clothes is the most tedious household chore. It's physically and mentally exhausting (unlike doing the dishes, which is only physically exhausting as I can wash piles of dishes while zoning out). And not to mention take so much time, a week's worth of unfolded pile of clothes for our entire household can take me 1-2 hours to fold uninterrupted (2-3 hours if I do it while watching Netflix or something).
Small sharing tips, in case it can help: have the maximum of hangers possible. It take a bit more space, but the time saved is huge + it remove iron time for low wrinkle sensitive clothes (e.g cotton polo)
Yep.. and with 3 kids so many of the time-consuming "tasks" are not open to automation because a human doing it is the point.. eg. bedtime books, bathing the kids, helping them make their lunch/breakfast, etc.
Sorting and folding laundry is one of the most tedious chores in our household. Do you have kids?
FYI I do have 3 kids, no, it doesn't take that long (we share this with the wife), in a few years the kids will be able to help us with this task so it's gonna be even less time. Doing household chores is good for the kids. I wouldn't wan't this robot driving around my household when I have small children anyway...
If I did have kids, I probably wouldn’t want them to run around a big heavy robot with very little spatial awareness.
Come back when you have a family and you run a washing machine full of clothes per day.
Don't you just throw all the washing into the washing machine, wait for it to actually do the work, and then hang it all up afterwards? It's like a 15 minute job max.
I mean folding after it has been dried. I hate folding clothes. In my case luckily my wife doesn't mind (as I don't mind doing the rest of the process) but if in say 25 years from now there will be a relatively small and affordable folding appliance, I would totally buy it.
Hopefully it won't take 25!
I don't know, was already bitten by this in ~2015 with self-driving cars and thinking that my kids will never ever need a driving license so now I'm really cautious on these topics.
How much self driving has arrived for you?
Not enough to think that in 8 years there will be L5 autonomous cars available to be bought (at a reasonable price) where I live.
L5 is impracticality high. The SAE levels aren't the right ones anyway. Comma.ai is only level 2 on their scale but it makes driving so much easier.
Yes, but anything outside L4 or L5 will still need a driving license to drive the car. In 2015 I though that around 2035 that would be the case, but now I doubt it seriously.
80% will be in the laundry bin, your first task before doing anything is either shouting around the house to bring in the other 80% right now or go on expedition yourself.
10 minutes later you are at the washing machine, where you'll never be able to fit all laundry so it's two runs minimum. Somebody will hopefully in advance, but probably only prompted by your call around the house mention a number of particular items will be needed later today, certainly before the second load will be dry. You are now searching for those items. Or arguing for why you really don't have time and parents are humans too and not unpaid servants.
Another 10 minutes later the first load is washing. You set a timer, or perhaps have a machine that pings you when done. Of course, you have More Shit To Do, so you don't notice your alarm or ping and by the end of the morning when you hoped to sit down to enjoy maybe a whole 15 minutes of reading move the first load to the dryer and put the second load on. You discover that somebody never emptied the dryer. Finders keepers!
1,5hr later you fortunately do not miss the alarm and take our the dry but not 100% dry laundry and move the second load to dry.
3hr later you empty the second dryer load, because the run took longer because nobody ever fucking cleans the lint-container. At least you had time to fold (which to the many item-types in your household and your poor fine motor control certainly does not take you _just_ 15 minutes). A cat decides to kick over a pile and leave hair in a second pile, causing you to redo some work. 30 minutes in your are done and finally have time to fold the second driers worth.
Do you get it now?
I replied to another comment, but I do have 3 kids.
Everybody's different, I guess, but for me this is absolutely a chore I would automate away, and I don't want to "force" another human being to do it for me.
I think its a ubiquitous task which requires some amount of precision dexterity, so its an easy to digest example.
A lot of smaller chores look trivial when considered in isolation, but have disproportionate time and mental cost because of how they fit in the flow of life. Sometimes the effect is similar to interrupting an engineer for a couple of minutes.
Funny you mentioned H&M. They are boycotted on China due to their public stance against slave labour:
https://www.bbc.com/news/business-57691415
It is important to remember that China is not only about open weight LLMs, fastest cars and flying taxis. There is still a lot of very serious human right abuse going on in this country.
I'm not super sure how relevant this is to the overall topic and thread, tbh. There's plenty of cheap yet real rebuttals, whether it's all the things with ICE (and not just the top-of-mind stuff; but effective deprivation of due process for suspects; etc), or even war crimes.
I don't think it serves much purpose tbh -- it's not going to change anyone's mind.
This was not a China/US good or bad comment. And I admit it was highly off topic.
But maybe once in a while we need to remind ourselves that the world is more than stock prices and LM arena benchmarks?
Yeah, western lecture on human rights carry no weight these days lol (Israel support, Trump doing Trump thing, war in Iran, etc). They ruined their own reputation so the rest of the world just tuned out.
When the last day's news were about other people shot dead by ICE without a plausible reason, it is hard to take seriously discussions about human rights abuse coming from USA. It is hard to imagine a more serious abuse than being randomly killed on the street, while walking or driving during your normal daily routine.
Unfortunately, there are very few, if any, countries remaining where human rights abuses are not frequent, so mentioning them in unrelated technology discussions has become non-relevant.
It's one of the most annoying.
I have a dishwasher so I do minimal washing up. I don't have a dryer as I air-dry my laundry but I could get one if I wanted to. I have a Roborock although I don't use it much since having a toddler.
But folding laundry is the one chore which still can't be automated. Even cooking itself has become much easier with air fryers etc.
I have a pile of washed clothes that moves from the bed to the chair for a good week before it finally goes into the closet.
It's meant for war. Today it folds your clothes, tomorrow it plants mines and shoots at you on a battlefield.
So is a pasta factory. Today it rolls and forms macaroni, tomorrow it makes bullets. Have you ever wondered why rigatoni are so wide? That's why.
The pasta factory analogy is hilarious, get an upvote. Is it something from the WWII?
No, not really. The macaroni are formed by squeezing the dough through the openings, while the cartridge cases are made by stretching small metallic disks into cylindrical shape (and then stamping the bottom to make it flat and have a proper rim), so it's impossible to repurpose the pasta-forming machines into the casing-making ones.
But it is an urban legend of sorts, along with "baby strollers with chain tracks instead of wheels", and shows how little most of us know about industrial manufacturing.
Seems overly complicated for that task (planting mines / shooting) when kamikaze drones are cheap and mass produced.
What?
There's already a better system today: mine launchers (remote minelaying system)
Perhaps what impressed me even more is how functional the page was without JS enabled!
"Did you pack the cases yourself sir?"
Please keep in mind that "uncut footage of ..." is the same as those instagram videos where people throw a basketball backwards 20 meters while drinking beer, straight into the goal.
It doesn't prove this is average performance. In fact, if it was, that's not what they'd be showing.
From my skim of their paper, they have gotten robodojo 14% versus previous sota 9%. So its an improvement, but doesnt seem like a reliable model yet, which is fine. I also notice the model becomes worse with more data on the printer refilling task, so thats either a bad run or its a data point against the robustness. The transfer learning results are meaningful. 75% on new complex tasks from <10 hours of demo is really good. I believe this is a 2x improvement over the SOTA. 10 trials per task is standard for these papers but is pretty weak statistically. Memory tasks its weaker at but they have acknowledged it, and I cant see why this arch would have problems w that in the future.
The fact it didn't do the 2 second tshirt fold trick shows it's just copying trained movements.
This is Automation not AI.
Well, couldn't it have been trained on the 2-second fold trick? Besides, it might require more precise or quick movements than a regular fold.
The movements are simple, but you must start with a much more precise grip, which seems difficult to achieve for a robot.
it looks like the robots are not just autonomous but also they do cooperate with each other.
pi.website
DO SOMETHING
What if mum but venture backed lmao
It makes me think of why couldnt they just automate ironing and folding. I can stick stuff in the suitcase myself.
You could get wrinkle free shirts and a plastic folding board / helper, and you are 90% there for 1% of the price ;-)
Automated ironing sounds like massive fire hazard even for showing off video.
Hand held steamer > ironing
The video cuts to a new shot every couple of second instead of one continuous shot like all robotics research. This makes it hard to trust it. The robot probably failed a million team and they stitched the video from 100 experiments.
The video also tells you that there is an uncut video at the bottom of the page. Did you take a look at that?
Nope did not notice. Updating my comment
The uncut video is at the bottom of the page.
And it shows the robot getting into something of a loop when trying to fold the white shorts. I wonder if someone pushed a button to get it out of that.