I can't wait to see the day I have to listen to the ad pitch of my robot butler before it goes make me a coffee.

The ads will run on the coffee cup which will be the only thing the robot will make coffee in.

And it will have a Keep Warm option for monthly subscribers.

I hope early home robotics will be like a game console launch - a community rush to break through any initial barriers to homebrew. Specific early firmwares will be desirable and a whole community will form around creating exploits and running user software. As old robot units fall out of service and are sold in bulk, the community will pick them up and keep them running.

which means that late stage robotics (or game consoles) become very heavily locked walled gardens as every exploit gets patched.

Sex robots.

I for one eagerly look forward to alien-autopsy like videos of people reverse engineering robots that they acquired by 'disabling' whatever equivalent of Lime/Bird/Volt service sends out on the streets to do stuff for people.

It'll be neat to walk through some weirdo mechanics shop that's full of robots in different states of disassembly that have been repurposed to help with whatever mad scientist hacker schemes that they have in mind.

Dont forget the artificial people personalities

And you'll better start to practice worshipping those ads unless your ass wants to be the star in B-166ER's last home video.

We are going to be conditioned into thinking 20k-50k for a robot is ok to spend. The question is how long before get there

Well yeah, if it can do most of the household chores and lasts a decade, it’s totally worth as much as a car to many people.

And we are being conditioned to think chores like getting drinks and putting the dishes away are beneath us. It is up there with motion- or voice-activated room lights. Want a robot to bring you coffee in bed each morning? Unless you are in a wheelchair, put the coffee machine on a timer and walk to the kitchen you lazy piece of ...

Dont know how to make toast? Too lazy to clean your own sheets? Everyone wanting one of these robots should have to do a few weeks of army basic training before earning the right to be this lazy. (Actuallty, iirc, we didnt wash our own sheets. But we did make our own toast!)

I agree there's a path where this encourages people to be even more sedentary and lazy. On the other hand, if I cook every meal and clean up after, plus take care of home cleaning, I could easily spend 3 hours a day on a variety of home tasks. A robot could potentially prepare better quality meals from fresh ingredients and save me 3 hours a day. It could also fix holes in my clothes and do other tasks I'm just not motivated enough to do. So the way to think about it is like you just gained a huge amount of energy and free time to do things you weren't doing before.

If it provides similar utility as a car, what's the big deal about that price? It would be the same type of purchase, right?

No. A car empowers me to do things i cannot do myself. That is very different than taking on mundane tasks to free up more time for more minecraft.

Or kids. Or work.

Nobody has to condition me to see a useful robot as being worth 20K+. Being able to offload domestic labor would be a huge win. So would the more subtle benefits of living in a home that is always in pristine condition.

Once they're able to cook, they'll be in every middle-class household on earth.

That robot will of course be transmitting 24/7 video and audio to the mothership and be fully remotely controllable, and its corporate overlord will be able to override any programming you give it. Of course, "to enhance your customer experience".

No amount of money in the world would convince me to allow such a monster in my house.

LG group uses it naturally for spying from data aggregated in home appliances. The rationale is fluff.

A humanoid robot would demand continuous maintenance, especially after planned obsolescence kicks in. No robot has ever worked under dirt conditions.

It is good to see someone finally working on one of these without the stupid legs

(I think the legs are stupid because they give your robot the chance to fall over and are not relevant for most environments).

Do you have stairs in your house?

I am protected

Goons never die, they just go to HN and regroup

I'm glad someone caught this.

Two robots

This is how I used to solve the roomba-can’t-use-stairs problem.

I’ve now moved to a single floor. Problem solved!

Then they can't bring anything up or down the stairs.

Still, I expect it won't matter - by the time we have reasonably priced robots that can reliably do all housework, that's like 90% of jobs eliminated from society and probably society will collapse.

Maybe one can throw things to the other :).

My kitchen, office, living room, gaming computer, entryway, and bedroom are all separated by sets of stairs.

So that's six robots for me?

To be fair, there are wheel arrangements which can negotiate stairs. Tank treads can probably do it too.

The problems with legs are sufficient that having two robots would be better than having one with legs. Here's why:

- You can increase the payload a lot

- You can increase the torso weight which means you can use cheaper, heavier components

- It won't fall down and hurt someone if it malfunctions

- It will never be in a position where it might fall down the stairs (if it were, it could kill someone)

- It will use less power/have a longer battery life

A carrier bot can be parked near the stairs to transport the bot to different floors

why not hyundai (given the existing boston dynamics ownership)?

Why not both? They partner with Anthropic, OpenAI, Google, etc.

So every 1000 physical bots are going to need the power and cooling equivalent of those super-popular datacenters they are trying to build everywhere

And I am virtually positive I know what the most popular application for them is going to become, the same way we've somehow decided to legalize gambling and drugs and make them available at a click to everyone everywhere

BTW are people going to be able to hack them to commit crimes? Protest for them?

But eventually everything is used for war to murder undesirables, we're only a decade away from the US or Israel etc. airdropping 1000 armed humanoid robots into a civilian space to hunt for "terrorists"

>we're only a decade away from the US or Israel etc. airdropping 1000 armed humanoid robots into a civilian space to hunt for "terrorists"

Currently they just bomb the buildings into the ground, killing everyone indiscriminately, so robots can't be any worse than that.

Nvidia needs to keep the hype train going, by throwing darts in every possible direction.

They’ve been throwing darts in this particular direction for many, many years - back when I was watching 2 minute papers with interest and hoovering up their stock at $0.80, long before the hype train pulled out of the station.

This is all mature stack, and the value is enormous and largely unrealised at this point - the systems they have for everything from training to implementation to edge inference basically present a complete capture of the ecosystem for robotics, autonomous vehicles, manufacturing, you name it - anything where you’re integrating input streams and acting on it, they’ve got covered, end to end.

I continue to hold a substantial chunk of Nvidia - because while hyperscaler spend may wane once the initial arms race subsides, they are uniquely positioned to pivot to making use of the output of hyperscaler and other GPU product.

There is no good reason to have a humanoid robot. None. Dishwashers do best in a dishwasher shape.

A couple weeks ago I saw a proud post by a humanoid robot CEO showing off how well his robots could sort objects on an assembly line. It was painful to watch. It's like he's never seen a real industrial sorting system in his life. Vision algos from 30 years ago along with simple mechanics or puffs of air were doing it 100x faster back then. I'm not convinced there is a bubble in AI, but there's definitely one in humanoid robots.

Yeah but the idea is that you can have a generalist robot that can do many different assembly line tasks.

There are plenty of manufacturing tasks that are still done by humans because it's too much hassle to make a dedicated robot to do it. Even on high volume car manufacturing it's very common to have human steps.

Sorting is just where they've got to so far; not the final destination.

This is the figure tweet?

Yeah I think it was

A robot that cleans my toilet would be nice. Sure I could have a dedicated toilet cleaning robot, and a dedicated dishwasher loading robot, and a dedicated pickup-up-crap-off-the-floor-the-kids-have-left-around-robot ... or just something general purpose?

I don't care if it is humanoid or not, but given that our house is built for humans to interact with it seems reasonable that it should fit into that space.

A robot with enough arm strength to properly clean a stubborn toilet wouldn't be safe around kids.

I let my current cleaner be around my kids.

Now the question is is it riskier to have basically a stranger with strong arms in my house near my kids, or a robot with strong arms in my house near my kids?

I feel like a robot has the technical capacity to see behind it and stop (I have many times for example been using the vacuum and moving my arm forwards and backwards and whacked a kid in the face with my elbow on the backswing because they've walked up behind me and I've not known, but a robot with literal eyes and radar in the back of its head would spot that situation and freeze). Similar to self-driving cars: they have lots more eyes than a human has, and can be looking everywhere at once etc.

But do we trust the programming? Do we trust the human cleaning my toilet's "programming" (thoughts, emotions, motives etc)?

Your cleaner is a human and you trust them to behave like a human including having a quite low probability of suddenly having a seizure. Robots do not think the same way. Any software glitch may cause it to move in an unexpected way. I had one freeze for half a second (unknown reason) while motors were full forward, and it rammed itself into a literal brick wall.

It just needs the patience to apply and gently scrub with mild chemicals for several hours.

You know what does this better than a humanoid robot? A toilet with a built-in dispenser for mild chemicals! It could do it periodically, or with every flush so that the mild cleaning chemicals are always sitting there. An unsophisticated one made of cheap plastic could be bought for $5 a the grocery store and clip onto the side.

A dishwasher is the only reason you can think of a humanoid robot? How about a robot to load and unload the dishwasher.

The fact is we live in a world built for humans. I have a robot vacuum and for it to be effective I had to setup my home in a certain way, and even then it is not fully effective.

People pay for cleaners to come into their home all the time, it shouldn't be hard to think why a humanoid robot would (theoretically, if it worked well) be far better than a purpose built machine in the home. But also in many cases working with those machines.

You could build fully humanoid robots that solve every problem generically, or you could solve the sub problems much more efficiently - I don't know why everyone is all in on the hardest problem when in software and hardware we find that the "solve everything" tool basically doesn't work, because solving everything is usually solving nothing very fast.

I don't think most people would want dozens of task specific robots in their house it would be far more economic to have a single general purpose robot that can use pre existing tools and methods.

Right, I am saying that the fundamental design of household objects need to change to allow much simpler maintenance without humanoid robots.

I think its easier to build a dish washer that can stack plates from first principles than humanoid robots. The cultural shift is the harder part.

You can buy a camera, portable music player, handheld gaming console, eBook reader, dumb phone, calculator, etc to solve each sub problem efficiently. Or you can buy a smart phone that generically solves every relevant problem. What direction did you take and how is life with it?

Those are all digital bit operating items that can be put into one box, the physical world doesn't work like that.

What makes the physical world different in this context? Isn't it things that humans are currently doing that we're automating away? Why not a human-like (in terms of capabilities) machine so it has similar flexibility?

Atoms are harder to reorganize than bits, undo is not an option because of entropy, game over means you are dead.

Yes, and my point is that human like is so difficult that solving lesser problems is a much more tractable approach.

How about integrating the cleaning mechanism into the dish cupboard so they don't need to be moved at all?

This exists. It's called dish drawers. Two mini-dishwashers in a unit with the idea that you will take your dishes out of one, use them, and put them in the other. When you need to, you run the dishwashing cycle in the dirty drawer. It does seem a little silly, but isn't storing your dishes directly in the dishwasher far more efficient than either manual or automated unloading?

Where is my non-humanoid robot that does all the household cleaning? Including vacuuming the stairs (a roomba doesn't work there), dusting the surfaces, mopping floors, cleaning windows...

I guess a warehouse can be designed in a way that works well for a non-humanoid robot, but an environment designed for people in the first place (like a home) fundamentally needs to be person-shaped.

Dishwashers also cannot work without a human loading and unloading dishes. How do you propose to automate this part?

It would actually be pretty easy to do with regular robotics (well, compared to a humanoid robot anyways). The reason nobody does it is that it takes up a ton of space which is at a huge premium in pretty much every kitchen ever.

Also like, loading and unloading the dishwasher is not that hard or time consuming.

> Also like, loading and unloading the dishwasher is not that hard or time consuming.

For a me a robot to do the dishwasher would be the number 1 reason for me to buy one.

My dishwasher is basically going at least twice sometimes three times a day (household with small kids). If I "miss" a slot to get everything washed before the next meal time then two things happen:

- the unwashed things begin to build up so there are too many things to fit in the next round and its hard to catch up.

- the things to need to use for meal-Y were still dirty from meal-X so you cant use foo etc.

Its "not much effort" true - perhaps 10-15 mins to unload then reload, but you need to do it 3 or 4 or more times a day AND you need to be there to do it on time so that there is time for it to finish it's load before meal-X etc.

If you are exhausted and its already 11pm and you've got to do your 3rd go at the dishwasher for the day so dirty things from dinner are getting washed and things are put away and ready for breakfast in the morning etc its really annoying. Its the last thing you want to do before going to bed. Or its morning and you're trying to get everyone out the door to school/work and the like, and you need to get the dishes going so that they're clean and ready to unload at lunch time (so that you can get the dirty lunch dishes in at lunch time etc).... you can see how this builds up into quite a pain in the ass hamster wheel.

I would 110% buy a humanoid robot for the cost of a decent second-hand car (so lets say about GBP10-15K) that was able to reliably do three or four 1 hour shifts per day doing basic house-keeper duties autonomously. So aforementioned dishes, cleaning down the dinner table, wiping down the kitchen worktops/countertops, picking up toys and cushions and shoes etc, then it can just go fold itself back into a cupboard in the kitchen to recharge for its next shift. Doesn't have to cook or play the violin or anything, basically just pick up crap off the floor and do the dishes every few hours so I don't have to. Bonus points if it can do it while I am working and/or it can do it silently at night

A man can dream.

You have a house with small kids and you want a 600 pound autonomous robot walking around? These aren’t funny tech products, they’re industrial equipment.

Figure 03 is 134 pounds. 1x NEO (the only one really designed for household usage) is 66 pounds. Unitree R1 is 55 pounds. None of these are ready for real work yet but probably at some point in the future we will have practical humanoids.

All of these are more than enough weight to kill a toddler by falling on them. That, of course, ignores all the other hazards presented by (presumably) titanium limbs attached to actuators able to exert enough force to crush any extremity you care to name.

The first person who has their child injured by one of these things will have a hell of a lawsuit on their hands.

A human on average weighs anywhere from ~110 - ~250 pounds (very broad guess). A human can also fall on and injure a child. And you can be sure that, similar to how bones are covered by layers of tissue and skin, robots designated for household use will be built with materials that improve safety during physical interaction. There isn't much difference.

Also the customer will very likely be asked to sign damage waivers and whatnot as part of the sale/rental agreement.

I think this is a good point, but it ignores the idea that a human form is not the ideal (or even closet to) one to do specific and generalized jobs.

Maybe not human form, but my house is designed for humans. It needs to be able to handle stairs. It needs to handle my doors and halls. It needs to reach upper shelves that are human height (my wife is short and constantly complains about things on a top shelf she cannot reach). I have corners in my house that a large dog can only manage because their spine bends. Human form in many ways is a constraint.

It is ideal insofar as the world has been created with humans in mind. And it’s easier to acquire training data for human shaped entities than for some hypothetical better shape.

I'd love the robot to fill and empty the dishwasher and put the stuff in the correct drawers and cabinets

edit: but if the robot could in addition also do dishes in the sink and not need a dishwasher at all, that'd also save up space in the kitchen for something else

I'd rather have two dishwashers and an automated loading/dispensing system built into each one. The dishwasher is already a fairly optimal dish storage device. Using somewhat standardized dish dimensions would make it fairly easy to implement.

I wish my kitchen was that big, but even then I think I'd prefer to have my plates stacked. Sure the ones my family uses every meal may as well live in the dishwasher, but I have a few more because once in a while I have guests (or they break) and I don't want those to have the empty space between them the dishwasher needed.

This exists.

Depends on what your definition of "good" is.

If your definition is "it could, at some point, enable me to stop paying humans for their labor and pass along more of the value to major shareholders like myself", then yes, that's a reason to want humanoid robots.

If your definition of "good" is a little more broadly scoped than the above - which it should be if you don't have an MBA and a substance abuse problem - then you're correct.

In a competitive market, this won't happen: "passing along more of the value to major shareholders like myself." The price of human labor will go down, but competition will force the price of goods to go down alongside it. Profit margins will stabilize, but the cost of living and the cost of goods will plummet. It's like the invention of the power loom: it was terrible for the wages of hand-weavers, but it made clothing radically cheaper and more abundant for the rest of humanity. The only way the shareholders keep all the value is if we allow monopolies to form.

The potential difference here is that it might eliminate all human labor which would likely force us into some new kind of economy. Hopefully something better than one where humans waste their lives on manual labor.

I found Manna to be a decent read, with its dual vision of how a post-human-labour world could be.

http://marshallbrain.com/manna1.htm

I'd love to see the cost of living go down. Unemployment raises fast; the cost of living goes down very slow, and it will take a few generations. That's why the underclass will go through a lot of pain.

More than half of the cost of living is land rent, which has nothing to do with human labor or cost of anything. It's a rather artificial forced payment that only increases when everything else gets cheaper.

There won't be any value at all to pass along, as wage labour is the base requirement for a capital-driven economy.

In the long-term, you're absolutely correct.

But the kind of person I'm describing doesn't care about the long-term, or really, anything beyond the current fiscal reporting period.

If there really were some sort of AI that were to be able to drive a humanoid robot to do the same tasks as well or better than a human, and we saw mass adoption, it'd take longer than a fiscal quarter to see the full macroeconomic impact, but the layoffs? You can write those down right then and there.

I have written this before, It is not AI with digital automation that worries me, generally speaking I think that is a good direction even thought we are still in early stage.

It is Humanoid, that will change everything. While we are still someway off, if we had PC - > Internet > Smartphone > AI, what is after AI will be Humanoid.

We still have another 4-5 years to go on current AI, and then Humanoid will further carry AI forward. This is similar to how Smartphone made the whole internet population 5 - 10 times bigger, further increasing demand on internet infrastructure. If anything, Apple should work on this. Perhaps the only thing that will be bigger than iPhone.

You say it worries you more than pure digital AI but don't explain. What is it about that humanoids that makes you worry?

I was thinking the other day that Apple is best positioned for drone companions- not even necessarily humanoid, but solving the battery charge problem to have an assistant follow you everywhere.

You’d only have to worry if we still had a long horizon of peace and globalism. With more potential breaks in supply chains, everything will cost a lot more in terms of time and money.

How did this "Physical AI" term originate? Isn't "robot" good enough anymore? Or "android" (though maybe that should be reserved for the really human-like machines)?

I recently published an article on this topic titled The Next Big Leap: https://arkvis.com/blog/2026-06-05_the-next-big-leap.html

Does this mean Nvidia actually has an actual working operating system? If so, what are they waiting for? And why do they need Microsoft? For their latest Arm venture.