6 ms·

The Coming Robot Home Invasion

29 points by walterbell 1y ago

alecco 1y ago
https://paulgraham.com/submarine.html https://paulgraham.com/submarine.html
- smallmancontrov 1y ago
  Just imagine how many suits people will buy for their robot butlers! Suits to the moon!
- zoeysmithe 1y ago
  I think this is a little cynical. Robotics and AI are moving at a decent pace. I just bought a robot vacuum that is also a mop and has all manner of sensors and even mapped my house. It was like $250. That's a huge leap over the $500 one I bought years ago that just bounced around randomly and only vacuumed. "Hey why doesnt this have an app and complex sensors and be able to dynamically make a map," would be borderline-laughable not that long ago, but now its the bare minimum for a robot vacuum. I don't even see a lot that don't also mop too. When before the mop feature was seen as this high-end feature, often requiring buying a 2nd machine. Now its just something tacked on. I'm not going to go into the controversial world of assisted and some level of self driving, but its clear that's making a lot of progress and today is just seen as a boring feature on your car, like seatbelts or ABS. I've been watching videos from Chinese robot manufacturers and we definitely are 'getting there.' What getting there means is the big question, but this isn't Omni magazine in 1985. You can see videos of these things doing a lot right now, even if its rudimentary. Unitree especially, considering how they aim for a 'low' price point. Of course a lot of these are carefully staged demos, but the fundamentals are here. Even non-staged stuff is starting to look relatively good. The recent leak about Apple's robot is interesting too. Apple is very conservative, so if they're testing one, even if its a tabletop (with an arm?) that says a lot on what's practically possible and affordable right now. The same way when Apple was testing the PR waters for an electric car, it was a great time to move into electric cars.
  - walterbell 1y ago
    > robot vacuum that is also a mop and has all manner of sensors and even mapped my house. It was like $250 Since we're already in the "submarine" sub-thread, make/model please :)
    - 0xffff2 1y ago
      I do would really like to know that make and if/how well the thing actually works. $250 is shockingly cheap compared to any iRobot product. If people are selling a $250 competitor that actually works, how on earth is iRobot even still in business?
      - Izkata 1y ago
        Brand recognition. About two or three years ago my old robot vacuum was finally dying and I did a lot of on and off searching for what was out there. When I started looking almost every competitor was better quality and features than what iRobot had, for around the same price, and they were just starting to play catch-up. I don't know how much they've improved in the past year, though.
      - baumandm 1y ago
        They might not be for long: https://www.theverge.com/news/757434/irobot-earnings-show-roomba-maker-still-struggling https://www.theverge.com/news/757434/irobot-earnings-show-ro...
    - vizzier 1y ago
      Not parent, but I've had good success from our Roborock Q5pro. Previously had a Ecovacs T9 aivi. The Ecovacs died twice, each just after a year (ie just out of warranty) needing to have main board/motor replaced. The bells and whistles didn't seem to be worth it. Self emptying station barely worked in an apartment with two people with long hair. Needed a lot of disposable replacement parts (main brush, side brushes, filters). Mopping feels like a bit of a gimmick. The much cheaper Roborock so far has been solid but we've only had it for 4 months. Mapping is marginally better (it is FAR less likely to reset the map on you which is a real annoyance). Filter is washable, main brushes need to be replaced less often. Has a better mechanism for collecting long hair off the brushes. Mop attachment still feels like a gimmick.
ikesau 1y ago
If I were to spend $10,000 on cleaning, I think I'd prefer to have 100 weeks of a human cleaner that can vacuum, scrub, climb stairs, etc. But, I take the point that more and more of these tasks will be automated more effectively in the coming decade.
- bluGill 1y ago
  Putting away toys is the hard part. Though as I get older my back is starting to demand help with some of the rest.
- simultsop 1y ago
  Isn't the plan of automation to let humans free at any cost? Like it happened with computers and relief from bureaucracy. First printers costed 20k, still there are 20k printers but way more powerful, the first series now are in homes under $100. It will take time to get cheap.
  - nightski 1y ago
    I don't think a rational market would want anything at "any" cost. Either the robot has to be cheaper, or better enough to justify the increased cost.
  - newsclues 1y ago
    Automation won’t make all humans free. It will make it easier for some humans to free of interaction with other humans.
- newsclues 1y ago
  I have a job as a cleaner. We have clients who spend about that hiring us to do regular weekly cleaning. But they still own a robotic vacuum, which they can deploy the other six days a week we aren’t there. Initially the people who buy home robots can afford humans and robots. Still so far away from being able to replace humans for all tasks, the difference between low hanging fruit (vacuum hardwood floors) and difficult tasks (dusting fragile art) is vast.
- 1y ago
1970-01-01 1y ago
I don't see it happening anytime soon. Maybe at $5k (in today dollars) in ten years it will be something, but not today. Boston Dynamics is arguably the world leader, and they are focused on industrial simply because "Rosie the Robot" just isn't capable of doing enough for the price point.
- fragmede 1y ago
  With all the competition coming out of China, Boston Dynamics is arguably not. They came out with videos of a robot years ago, and are selling a $70k robot dog platform years, but Unitree is shipping a humanoid robot today.
brk 1y ago
This is not going to happen any time soon. The acquisition cost and ongoing maintenance issues alone limit the practical market of home robots to a tiny segment of the market. As neat as the idea sounds, the practicalities and edge cases keep this in the science fiction category right now and for the foreseeable future.
- nebula8804 1y ago
  People have more money than sense so at the very least, it will be another piece of junk that accumulates in peoples garages but at the very best it could serve some niche for some people.
  - Joel_Mckay 1y ago
    Bipedal humanoid robot platforms are a terrible design. Even if the thing fell over it could inflict serious injury. =3
    - nebula8804 1y ago
      Have you seen that Chinese Robot walking around American Cities? [1]: https://www.youtube.com/shorts/UpXKZEgrB6Q https://www.youtube.com/shorts/UpXKZEgrB6Q There was a clip where it almost fell down a set of stairs but it caught itself. I cant seem to find it. There is this though: https://www.youtube.com/shorts/8Drm_v3_iG4 https://www.youtube.com/shorts/8Drm_v3_iG4
      - Joel_Mckay 1y ago
        Indeed, part of the design problem is the platform is inherently unstable, and energy inefficient. Giant robotic house spiders make far more sense. =3
      - Izkata 1y ago
        You can put the hash in a regular youtube URL to avoid the annoying Shorts interface and get back full functionality: https://www.youtube.com/watch?v=UpXKZEgrB6Q https://www.youtube.com/watch?v=UpXKZEgrB6Q https://www.youtube.com/watch?v=8Drm_v3_iG4 https://www.youtube.com/watch?v=8Drm_v3_iG4
everdrive 1y ago
I strongly believe that except in edge cases, people spend so much time cleaning because they're not thinking about process and lack attention to detail. (or, potentially, they have people in their family who lack these things) There are a number of "cleaning challenges" which are really just conscientiousness or process problems. "It's difficult to sweep the house" - Often, this just means that the house is too cluttered. You own too many things, they're not put away, and so each time you need to sweep the floors, you need to spend time organizing first. "The fridge is always getting dirty." - Only clean things should be put in the fridge. If your shelves are dirty, it's because the bottom of the items you're placing on your shelves are also dirty. Do you let ketchup run down the side of the bottle? Is your counter dirty, and you've set a clean bottle of salad dressing on a dirty counter? In both cases, you have transferred dirt and/or oil into the fridge on the shelf. And then this compounds multiple times: do you always put the ketchup in a different place? Well now you've spread around the dirt. Do you clean the fridge but bot the bottom of the ketchup bottle? Well, now your clean fridge is immediately dirty again the moment you reload it. - etc. So many "messes" in the house are compounding problems; the worse you let it get, the bigger the problem you've made for yourself. Didn't put away the kid's toys? Well now you can't sweep that area. Can't sweep that area? Well now you can't mop that area. etc. One solution, which people might not think about with regard to cleaning would actually just be to make sure the kid has fewer toys, or at least fewer _available_ toys. If you nail this, the clutter --> sweeping --> mopping cascade starts improving without any additional work on your end.
- AppleBananaPie 1y ago
  This is 100% describes my situation and I'm saved by the luxury of people coming to clean my home and reset it. Separately, I need to clean my fridge because then I can follow your great advice XD
  - hoss1474489 1y ago
    Make sure to clean everything that was inside the fridge before restocking or you reset your progress
- 1y ago
politician 1y ago
It would be amazing to have robot-delivered butler service. However, I'm deeply concerned about privacy and, given OTA updates, physical safety. There are enough bad actors on the Internet as-is, I don't want my butler deciding to reenact Order 66 after a Wifi update.
- Hizonner 1y ago
  It's worse than OTA updates. I'd expect most of these to use the cloud for a lot of their "thinking", and definitely to be controllable from the cloud on a minute-by-minute basis. Lose the WiFi and you can expect to lose a big chunk of the capability you paid for. ... and you can expect the subscription costs to keep going up. None of those OTA updates are going to give you any new value unless you pay more.
mattlondon 1y ago
I am fairly confident that while they're not ready "today", with the advances we're seeing in AI recently, we're going to have a viable in-home humanoid robot in the next decade or two, priced at "car" kinda prices I think this is going to be fairly transformational, at least for care scenarios. I'm not talking about the sci-fi pipe-dream of a Star Trek Data style robot, but more like a basic humanoid robot that can reliably do mundane & basic things on-command. Like pick things up off of the floor, go fetch items from another room, open the curtains, do some basic food preparation (e.g. heat things up in a microwave levels of sophistication, or even just getting a glass of water), do the dishwasher, take out the trash and so on. Even if it can only do 1hr of chores at a time before heading back to recharge, thats huge. It will help people live more independent lives for much longer before needing expensive care from humans (something that we have a bit of a ticking time-bomb of, at least in the UK where the population is aging rapidly). It doesn't need to be a fully fledged "robot-nurse" to be helpful. Bonus points if it can monitor it's user(s) and call for help if there appears to be anything wrong (e.g. fallen and can't get up type monitoring), or intervene in situations before they escalate (e.g. turn the gas off if it is left on, remind to take medicines etc)
- jppope 1y ago
  pretty sure we've already had products in the market that can do the things you are thinking of, the problem with products like baxter (or boston dynamics products) wasn't/isn't the capabilities it was/is the price point. the unit economics for robotics have always been the problem. Sure you can buy a robot for $45K but you can have a hell of a whole lot of days of cleaning with a cleaning service for that same money... the downstream effect of this is that it is very hard to get economies of scale so they never drop to the point where these things are normalized.
  - mattlondon 1y ago
    You can already buy one from unitree for USD16k. Full time care for the elderly is epically expensive. As the capabilities of humanoid robots grow, the demand and so the supply will increase.
mythrwy 1y ago
For this to work efficiently maybe we will need to evaluate the form factor of homes in general (particularly in the US). An example is olives. For centuries people picked or beat olives off of big trees but with machinery they have developed miniature olives that grow in neat hedgerows and can be easily harvested by machines. I keep thinking about how much space people actually need to live. Bedrooms in particular. If all you need a bedroom for is sleep or romantic activities, why do bedrooms need to be 200+ square feet? Clothes ideally should be stored in another place where you dress and that place should be close to where clothes are laundered and where you shower and prep for the day. An annex to the bathroom, a dressing room seems ideal to me for clothes storage. In this case I'd be perfectly fine with a capsule bedroom big enough to hold a bed, a nightstand and nothing else. I don't have any other concrete suggestions that would make automation easier but I'm sure this is possible.
- aetherson 1y ago
  I agree that you could imagine that homes could be laid out more efficiently for automation. However, the vast majority of housing stock is not new and is not cheap to remodel. The result is going to be incredible pressure to make robots fit existing houses, otherwise your market share is going to be like 2% of the TAM.
joshuaheard 1y ago
I live in a large suburban home. I spend $1,000/mo on landscaping services and $1,000/mo on maid services. If I could buy a $10,000 robot that could do those things, as well as have apps for doing dinner dishes, laundry, making the bed, and feeding the pets, I would do it in a second.
- nebula8804 1y ago
  You really sound like the exception even for this forum. Either that or I am woefully underpaid.
  - simantel 1y ago
    At $250/visit, that's just weekly landscaping and housecleaning. More than I pay for myself (we have a housecleaner that comes monthly), but not unreasonable, especially for a well-paid dual-income household. Re: Your compensation, Levels can be eye opening: https://www.levels.fyi/?compare=Facebook,Microsoft,Google,Amazon&track=Software%20Engineer https://www.levels.fyi/?compare=Facebook,Microsoft,Google,Am...
    - nebula8804 1y ago
      The person said $1,000 a month for landscaping and $1,000 a month for house cleaning. Thats like 1/3 of my take home. Is that really normal for people? Regarding salary, Looking at my area, unless im in FAANG seems like im making around what the site says. I feel like that expense would be too much but maybe I just didnt get into stocks or crypto as much as others have done here.
      - simantel 1y ago
        Yeah $1,000 a month would be $250 a week, which is 2-3 person hours of a housecleaner or landscaper's time, so like a 2 person crew for an hour or 90 minutes. If you and your spouse each make $150k (so $300k household income), paying 8% of your income ($24k/year) to have 6 extra hours a week to spend with your kids seems like a fair trade. Two people working in tech could easily be making more like a $5-600k household income though, at which point $24k is only 4-5%.
j_timberlake 1y ago
Absolutely not happening. If robots get versatile enough to clean a random home, then they'll be good enough for higher-value work like being a robotic soldier or building more robots. And if they're taking lots of jobs like that, then there's going to be tons of spare man-hours for busywork like cleaning.
- fragmede 1y ago
  That seems to say it'll eventually happen, the question is just on what timeframe. Three years? Five? Ten? Fifty?
groos 1y ago
We, in the tech industry, go all starry-eyed when we imagine utopian futures filled with robots that will do all your tedious chores. But a dark side of this that goes unnoticed is that such automation is often just a way for money to be transferred from the poorest sections of society to the richest. This fact was driven home to me recently when dropping my mother off at the airport. For various reasons, my mother us to use wheelchair assistance when flying, and I usually get a "gate pass" to accompany her to the gate. An airline provided wheelchair is pushed by an airline employee whom I usually tip a decent amount every time. The last time she flew, the employee insisted that we had to make a detour, otherwise they would "write him up". Not wishing to make his life hard, we agreed and he brought us to an area where they were trialing robotic wheelchairs. They insisted my mother use the robotic wheelchair and since I was with her an a little curious, we agreed. In the process of tranferring her, I forgot to tip the wheelchair pusher. The robotic wheelchair was an unconditional disaster. In a busy airport, it stopped every time it detected anyone within 5 feet of it and took 20 min to get where it would have taken less than 5 with the cherry on top being that it stopped and told my mother to get off 3 gates away. Had I not been with her, I'm not sure what she would have done, being too weak to lift her carry-on and walking to the gate. But thinking about it a bit, I realized that sooner or later, the robotic wheelchair would be improved to the point that it did just as good a job as the human pusher at which point, all the income earned by the humans, all of whom are from the poorer sections of society, would be transferred to the robotic wheelchair company and its shareholders, all of whom are probably in the top 2% wealth-wise. This transfer of wealth from the poor to the rich seems to be a recurrent theme in tech. Sure new wealth eventually is made, but it seems it's mostly for the top strata, with us techies being paid well enough to prevent us from feeling too many scruples.
cdcox 1y ago
The B2C case here seems off to me. The market of people who are going to pay the high price tag, have enough storage, and tolerate the extreme limitations and slowness of these machines seems to be: the rich, the elderly, and people with physical disabilities. The latter two categories come with a huge number of liability and regulatory costs that I think most of these companies are not willing to handle. That does not feel like a huge market. I'll have a stronger belief in these things for consumers when we start seeing B2B adoption. Hotels have routine, highly structured cleaning tasks; hospitals have a need for extra strength, have highly structured cleaning tasks, and need to stock items; grocery stores have highly structured cleaning tasks and need to stock. Hospitals at least could tolerate the slowness of these things. Without any B2B adoption it's hard to not see this as Roombas all over again. Cool for people who like it but low impact and still a toy 20 years later. I think generative AI makes these things better, though still perhaps struggles with long task adoption, but if you look at their movement they are still slow, weak, cognitively inflexible, and unstable. Maybe this tech is accelerating in some way I don't see and I'd love to be proven wrong here.
charlie90 1y ago
I don't get the naysayers, I absolutely see this happening. Take senior care, absurdly expensive even with low cost human labor, $100K a year. If we had $50k robots, it would be a game changer. People could live out their final years in their home with 24/7 personal care.
teekert 1y ago
But will Isaac have pain receptors? (Sorry, couldn’t resist this Orville reference)

29 points by walterbell 1y ago

alecco 1y ago

https://paulgraham.com/submarine.html https://paulgraham.com/submarine.html

smallmancontrov 1y ago
Just imagine how many suits people will buy for their robot butlers! Suits to the moon!
zoeysmithe 1y ago
I think this is a little cynical. Robotics and AI are moving at a decent pace. I just bought a robot vacuum that is also a mop and has all manner of sensors and even mapped my house. It was like $250. That's a huge leap over the $500 one I bought years ago that just bounced around randomly and only vacuumed. "Hey why doesnt this have an app and complex sensors and be able to dynamically make a map," would be borderline-laughable not that long ago, but now its the bare minimum for a robot vacuum. I don't even see a lot that don't also mop too. When before the mop feature was seen as this high-end feature, often requiring buying a 2nd machine. Now its just something tacked on. I'm not going to go into the controversial world of assisted and some level of self driving, but its clear that's making a lot of progress and today is just seen as a boring feature on your car, like seatbelts or ABS. I've been watching videos from Chinese robot manufacturers and we definitely are 'getting there.' What getting there means is the big question, but this isn't Omni magazine in 1985. You can see videos of these things doing a lot right now, even if its rudimentary. Unitree especially, considering how they aim for a 'low' price point. Of course a lot of these are carefully staged demos, but the fundamentals are here. Even non-staged stuff is starting to look relatively good. The recent leak about Apple's robot is interesting too. Apple is very conservative, so if they're testing one, even if its a tabletop (with an arm?) that says a lot on what's practically possible and affordable right now. The same way when Apple was testing the PR waters for an electric car, it was a great time to move into electric cars.
- walterbell 1y ago
  > robot vacuum that is also a mop and has all manner of sensors and even mapped my house. It was like $250 Since we're already in the "submarine" sub-thread, make/model please :)
  - 0xffff2 1y ago
    I do would really like to know that make and if/how well the thing actually works. $250 is shockingly cheap compared to any iRobot product. If people are selling a $250 competitor that actually works, how on earth is iRobot even still in business?
    - Izkata 1y ago
      Brand recognition. About two or three years ago my old robot vacuum was finally dying and I did a lot of on and off searching for what was out there. When I started looking almost every competitor was better quality and features than what iRobot had, for around the same price, and they were just starting to play catch-up. I don't know how much they've improved in the past year, though.
    - baumandm 1y ago
      They might not be for long: https://www.theverge.com/news/757434/irobot-earnings-show-roomba-maker-still-struggling https://www.theverge.com/news/757434/irobot-earnings-show-ro...
  - vizzier 1y ago
    Not parent, but I've had good success from our Roborock Q5pro. Previously had a Ecovacs T9 aivi. The Ecovacs died twice, each just after a year (ie just out of warranty) needing to have main board/motor replaced. The bells and whistles didn't seem to be worth it. Self emptying station barely worked in an apartment with two people with long hair. Needed a lot of disposable replacement parts (main brush, side brushes, filters). Mopping feels like a bit of a gimmick. The much cheaper Roborock so far has been solid but we've only had it for 4 months. Mapping is marginally better (it is FAR less likely to reset the map on you which is a real annoyance). Filter is washable, main brushes need to be replaced less often. Has a better mechanism for collecting long hair off the brushes. Mop attachment still feels like a gimmick.

ikesau 1y ago

If I were to spend $10,000 on cleaning, I think I'd prefer to have 100 weeks of a human cleaner that can vacuum, scrub, climb stairs, etc. But, I take the point that more and more of these tasks will be automated more effectively in the coming decade.

bluGill 1y ago
Putting away toys is the hard part. Though as I get older my back is starting to demand help with some of the rest.
simultsop 1y ago
Isn't the plan of automation to let humans free at any cost? Like it happened with computers and relief from bureaucracy. First printers costed 20k, still there are 20k printers but way more powerful, the first series now are in homes under $100. It will take time to get cheap.
- nightski 1y ago
  I don't think a rational market would want anything at "any" cost. Either the robot has to be cheaper, or better enough to justify the increased cost.
- newsclues 1y ago
  Automation won’t make all humans free. It will make it easier for some humans to free of interaction with other humans.
newsclues 1y ago
I have a job as a cleaner. We have clients who spend about that hiring us to do regular weekly cleaning. But they still own a robotic vacuum, which they can deploy the other six days a week we aren’t there. Initially the people who buy home robots can afford humans and robots. Still so far away from being able to replace humans for all tasks, the difference between low hanging fruit (vacuum hardwood floors) and difficult tasks (dusting fragile art) is vast.
1y ago

1970-01-01 1y ago

I don't see it happening anytime soon. Maybe at $5k (in today dollars) in ten years it will be something, but not today. Boston Dynamics is arguably the world leader, and they are focused on industrial simply because "Rosie the Robot" just isn't capable of doing enough for the price point.

fragmede 1y ago
With all the competition coming out of China, Boston Dynamics is arguably not. They came out with videos of a robot years ago, and are selling a $70k robot dog platform years, but Unitree is shipping a humanoid robot today.

brk 1y ago

This is not going to happen any time soon. The acquisition cost and ongoing maintenance issues alone limit the practical market of home robots to a tiny segment of the market. As neat as the idea sounds, the practicalities and edge cases keep this in the science fiction category right now and for the foreseeable future.

nebula8804 1y ago
People have more money than sense so at the very least, it will be another piece of junk that accumulates in peoples garages but at the very best it could serve some niche for some people.
- Joel_Mckay 1y ago
  Bipedal humanoid robot platforms are a terrible design. Even if the thing fell over it could inflict serious injury. =3
  - nebula8804 1y ago
    Have you seen that Chinese Robot walking around American Cities? [1]: https://www.youtube.com/shorts/UpXKZEgrB6Q https://www.youtube.com/shorts/UpXKZEgrB6Q There was a clip where it almost fell down a set of stairs but it caught itself. I cant seem to find it. There is this though: https://www.youtube.com/shorts/8Drm_v3_iG4 https://www.youtube.com/shorts/8Drm_v3_iG4
    - Joel_Mckay 1y ago
      Indeed, part of the design problem is the platform is inherently unstable, and energy inefficient. Giant robotic house spiders make far more sense. =3
    - Izkata 1y ago
      You can put the hash in a regular youtube URL to avoid the annoying Shorts interface and get back full functionality: https://www.youtube.com/watch?v=UpXKZEgrB6Q https://www.youtube.com/watch?v=UpXKZEgrB6Q https://www.youtube.com/watch?v=8Drm_v3_iG4 https://www.youtube.com/watch?v=8Drm_v3_iG4

everdrive 1y ago

I strongly believe that except in edge cases, people spend so much time cleaning because they're not thinking about process and lack attention to detail. (or, potentially, they have people in their family who lack these things) There are a number of "cleaning challenges" which are really just conscientiousness or process problems. "It's difficult to sweep the house" - Often, this just means that the house is too cluttered. You own too many things, they're not put away, and so each time you need to sweep the floors, you need to spend time organizing first. "The fridge is always getting dirty." - Only clean things should be put in the fridge. If your shelves are dirty, it's because the bottom of the items you're placing on your shelves are also dirty. Do you let ketchup run down the side of the bottle? Is your counter dirty, and you've set a clean bottle of salad dressing on a dirty counter? In both cases, you have transferred dirt and/or oil into the fridge on the shelf. And then this compounds multiple times: do you always put the ketchup in a different place? Well now you've spread around the dirt. Do you clean the fridge but bot the bottom of the ketchup bottle? Well, now your clean fridge is immediately dirty again the moment you reload it. - etc. So many "messes" in the house are compounding problems; the worse you let it get, the bigger the problem you've made for yourself. Didn't put away the kid's toys? Well now you can't sweep that area. Can't sweep that area? Well now you can't mop that area. etc. One solution, which people might not think about with regard to cleaning would actually just be to make sure the kid has fewer toys, or at least fewer _available_ toys. If you nail this, the clutter --> sweeping --> mopping cascade starts improving without any additional work on your end.

AppleBananaPie 1y ago
This is 100% describes my situation and I'm saved by the luxury of people coming to clean my home and reset it. Separately, I need to clean my fridge because then I can follow your great advice XD
- hoss1474489 1y ago
  Make sure to clean everything that was inside the fridge before restocking or you reset your progress
1y ago

politician 1y ago

It would be amazing to have robot-delivered butler service. However, I'm deeply concerned about privacy and, given OTA updates, physical safety. There are enough bad actors on the Internet as-is, I don't want my butler deciding to reenact Order 66 after a Wifi update.

Hizonner 1y ago
It's worse than OTA updates. I'd expect most of these to use the cloud for a lot of their "thinking", and definitely to be controllable from the cloud on a minute-by-minute basis. Lose the WiFi and you can expect to lose a big chunk of the capability you paid for. ... and you can expect the subscription costs to keep going up. None of those OTA updates are going to give you any new value unless you pay more.

mattlondon 1y ago

I am fairly confident that while they're not ready "today", with the advances we're seeing in AI recently, we're going to have a viable in-home humanoid robot in the next decade or two, priced at "car" kinda prices I think this is going to be fairly transformational, at least for care scenarios. I'm not talking about the sci-fi pipe-dream of a Star Trek Data style robot, but more like a basic humanoid robot that can reliably do mundane & basic things on-command. Like pick things up off of the floor, go fetch items from another room, open the curtains, do some basic food preparation (e.g. heat things up in a microwave levels of sophistication, or even just getting a glass of water), do the dishwasher, take out the trash and so on. Even if it can only do 1hr of chores at a time before heading back to recharge, thats huge. It will help people live more independent lives for much longer before needing expensive care from humans (something that we have a bit of a ticking time-bomb of, at least in the UK where the population is aging rapidly). It doesn't need to be a fully fledged "robot-nurse" to be helpful. Bonus points if it can monitor it's user(s) and call for help if there appears to be anything wrong (e.g. fallen and can't get up type monitoring), or intervene in situations before they escalate (e.g. turn the gas off if it is left on, remind to take medicines etc)

jppope 1y ago
pretty sure we've already had products in the market that can do the things you are thinking of, the problem with products like baxter (or boston dynamics products) wasn't/isn't the capabilities it was/is the price point. the unit economics for robotics have always been the problem. Sure you can buy a robot for $45K but you can have a hell of a whole lot of days of cleaning with a cleaning service for that same money... the downstream effect of this is that it is very hard to get economies of scale so they never drop to the point where these things are normalized.
- mattlondon 1y ago
  You can already buy one from unitree for USD16k. Full time care for the elderly is epically expensive. As the capabilities of humanoid robots grow, the demand and so the supply will increase.

mythrwy 1y ago

For this to work efficiently maybe we will need to evaluate the form factor of homes in general (particularly in the US). An example is olives. For centuries people picked or beat olives off of big trees but with machinery they have developed miniature olives that grow in neat hedgerows and can be easily harvested by machines. I keep thinking about how much space people actually need to live. Bedrooms in particular. If all you need a bedroom for is sleep or romantic activities, why do bedrooms need to be 200+ square feet? Clothes ideally should be stored in another place where you dress and that place should be close to where clothes are laundered and where you shower and prep for the day. An annex to the bathroom, a dressing room seems ideal to me for clothes storage. In this case I'd be perfectly fine with a capsule bedroom big enough to hold a bed, a nightstand and nothing else. I don't have any other concrete suggestions that would make automation easier but I'm sure this is possible.

aetherson 1y ago
I agree that you could imagine that homes could be laid out more efficiently for automation. However, the vast majority of housing stock is not new and is not cheap to remodel. The result is going to be incredible pressure to make robots fit existing houses, otherwise your market share is going to be like 2% of the TAM.

joshuaheard 1y ago

I live in a large suburban home. I spend $1,000/mo on landscaping services and $1,000/mo on maid services. If I could buy a $10,000 robot that could do those things, as well as have apps for doing dinner dishes, laundry, making the bed, and feeding the pets, I would do it in a second.

nebula8804 1y ago
You really sound like the exception even for this forum. Either that or I am woefully underpaid.
- simantel 1y ago
  At $250/visit, that's just weekly landscaping and housecleaning. More than I pay for myself (we have a housecleaner that comes monthly), but not unreasonable, especially for a well-paid dual-income household. Re: Your compensation, Levels can be eye opening: https://www.levels.fyi/?compare=Facebook,Microsoft,Google,Amazon&track=Software%20Engineer https://www.levels.fyi/?compare=Facebook,Microsoft,Google,Am...
  - nebula8804 1y ago
    The person said $1,000 a month for landscaping and $1,000 a month for house cleaning. Thats like 1/3 of my take home. Is that really normal for people? Regarding salary, Looking at my area, unless im in FAANG seems like im making around what the site says. I feel like that expense would be too much but maybe I just didnt get into stocks or crypto as much as others have done here.
    - simantel 1y ago
      Yeah $1,000 a month would be $250 a week, which is 2-3 person hours of a housecleaner or landscaper's time, so like a 2 person crew for an hour or 90 minutes. If you and your spouse each make $150k (so $300k household income), paying 8% of your income ($24k/year) to have 6 extra hours a week to spend with your kids seems like a fair trade. Two people working in tech could easily be making more like a $5-600k household income though, at which point $24k is only 4-5%.

j_timberlake 1y ago

Absolutely not happening. If robots get versatile enough to clean a random home, then they'll be good enough for higher-value work like being a robotic soldier or building more robots. And if they're taking lots of jobs like that, then there's going to be tons of spare man-hours for busywork like cleaning.

fragmede 1y ago
That seems to say it'll eventually happen, the question is just on what timeframe. Three years? Five? Ten? Fifty?

groos 1y ago

We, in the tech industry, go all starry-eyed when we imagine utopian futures filled with robots that will do all your tedious chores. But a dark side of this that goes unnoticed is that such automation is often just a way for money to be transferred from the poorest sections of society to the richest. This fact was driven home to me recently when dropping my mother off at the airport. For various reasons, my mother us to use wheelchair assistance when flying, and I usually get a "gate pass" to accompany her to the gate. An airline provided wheelchair is pushed by an airline employee whom I usually tip a decent amount every time. The last time she flew, the employee insisted that we had to make a detour, otherwise they would "write him up". Not wishing to make his life hard, we agreed and he brought us to an area where they were trialing robotic wheelchairs. They insisted my mother use the robotic wheelchair and since I was with her an a little curious, we agreed. In the process of tranferring her, I forgot to tip the wheelchair pusher. The robotic wheelchair was an unconditional disaster. In a busy airport, it stopped every time it detected anyone within 5 feet of it and took 20 min to get where it would have taken less than 5 with the cherry on top being that it stopped and told my mother to get off 3 gates away. Had I not been with her, I'm not sure what she would have done, being too weak to lift her carry-on and walking to the gate. But thinking about it a bit, I realized that sooner or later, the robotic wheelchair would be improved to the point that it did just as good a job as the human pusher at which point, all the income earned by the humans, all of whom are from the poorer sections of society, would be transferred to the robotic wheelchair company and its shareholders, all of whom are probably in the top 2% wealth-wise. This transfer of wealth from the poor to the rich seems to be a recurrent theme in tech. Sure new wealth eventually is made, but it seems it's mostly for the top strata, with us techies being paid well enough to prevent us from feeling too many scruples.

cdcox 1y ago

The B2C case here seems off to me. The market of people who are going to pay the high price tag, have enough storage, and tolerate the extreme limitations and slowness of these machines seems to be: the rich, the elderly, and people with physical disabilities. The latter two categories come with a huge number of liability and regulatory costs that I think most of these companies are not willing to handle. That does not feel like a huge market. I'll have a stronger belief in these things for consumers when we start seeing B2B adoption. Hotels have routine, highly structured cleaning tasks; hospitals have a need for extra strength, have highly structured cleaning tasks, and need to stock items; grocery stores have highly structured cleaning tasks and need to stock. Hospitals at least could tolerate the slowness of these things. Without any B2B adoption it's hard to not see this as Roombas all over again. Cool for people who like it but low impact and still a toy 20 years later. I think generative AI makes these things better, though still perhaps struggles with long task adoption, but if you look at their movement they are still slow, weak, cognitively inflexible, and unstable. Maybe this tech is accelerating in some way I don't see and I'd love to be proven wrong here.

charlie90 1y ago

I don't get the naysayers, I absolutely see this happening. Take senior care, absurdly expensive even with low cost human labor, $100K a year. If we had $50k robots, it would be a game changer. People could live out their final years in their home with 24/7 personal care.

teekert 1y ago

But will Isaac have pain receptors? (Sorry, couldn’t resist this Orville reference)