Hacker News Reader

Comment by SoftTalker | Hacker News Reader

Seems like a mistake. AI in its current form has limited usefulness for most people. Not something I would pay for to use outside of work. But a household robot that could clean, wash and fold the laundry, do the dishes, maybe even be a chauffeur... that would be huge. I think a lot of people would pay new-car money for something like that.

View on HN

> AI in its current form has limited usefulness for most people.

That's not what I'm seeing. My mom always wanted Google to just answer questions, and now ChatGPT can. She uses it enough in her daily life that she bought a subscription.

Yes, she knows it hallucinates and you have to double check everything, but so far she finds a ton of use for it even with those caveats.

Now, I agree that a personal servant robot would get a ton of business. Even at new-car prices, it's still cheaper than a human caretaker/maid/butler/etc. And the maid usually doesn't also mow the lawn on a hot day, while a robot would potentially do all kinds of different things without complaint.

> That's not what I'm seeing. My mom always wanted Google to just answer questions, and now ChatGPT can. She uses it enough in her daily life that she bought a subscription.

Reminds me a lot of AskJeeves :)

Google summaries from websites mainly did this before “AI” came around. Now I just don’t know the source unless I dig myself. Would rather it went back to before, or at least they implemented the veracity checks better.

I thought it was interested that the movie looper had argidrones, but what inspired that wasn’t part of the training set or discussion online. I get remarkably bad answers and search results which seem misleading at best.

> Now I just don’t know the source unless I dig myself.

Claude includes links to references.

What's a problem though is that while it includes a reference, its distillation from the source, or worse, from the combination of sources, is often plain wrong. I check them regularly, and it made me very distrustful of Claude's capacity to faithfully summarise or explain from a source.

When asked to give specific links, it's usually even worse.

I don’t disagree. Many times when I’ve checked the source links a language model provides, I get concerned. It’s definitely important to stay vigilant and exercise critical thinking, but the model’s ability to synthesize information and converse with it in natural language still blows my mind sometimes

So does ChatGPT. The hallucination problem is not completely solved, but much better than a few years ago, especially if you use reasoning mode, where it’s more likely to spontaneously do a web search

Those references often have no actual relation to the answer. A surprising amount of the time, if you check them they’ll prove it wrong.

I have a feeling that humanoid robots will have other, more intimate, tasks first. Our most primal drives seem to drive the advancement of technology.

Also, mowing the lawn on a very hot day is pretty bad for the grass.

I knew someone was gonna go there.

Yeah, I felt that someone had to speak up about the health of the lawn.

I think more non-tech workers are buying subscriptions than people in industry.

Neither I not almost all coworkers I know do not pay for AI subscriptions (or companies do).

But quite a few people’s sisters and parents and friends are paying $20-200/month for ai to help them with their “projects.” Whatever that means.

FWIW, Google does just answer questions now.

We already have robot grass trimmers and they work pretty well. Why would you want a crude, inefficient, facsimile of a human push or power an inefficient form of mowing the lawn?

Cameras worked pretty well, but most people take photos with their phones nowadays. Pro photographers still mostly use cameras though.

I can envision a future where people have a humanoid trimming a small backyard but at the same time the maintenance of e.g. a golf course would be done by dedicated robots.

EDIT: my point is that just like smartphones replaced a number of specialised devices for most people, a humanoid robot could do the same, by virtue of being a general purpose machine

I don’t want a crude facsimile.

I want a pretty good facsimile that can cook, clean, mow, carry stuff around the yard.

I want this much more than a robot lawnmower + robot vaccuum + plus other single purpose automatons.

For when they don't work pretty well and for a thousand other things besides trimming grasss.

Because houses and tools are made for humans.

You're honestly comparing a computer that can poorly answer questions to a fully functioning robot that does chores?

Well, one of those is already a reality in each persons pocket, and the other is a vision needing a lot of money and effort to implement at all

> robot maid that could clean, wash and fold the laundry, do the dishes, etc. would be huge. I think a lot of people would pay new-car money for something like that.

Once you take maintenance of a machine with price-parity to a new car into consideration, it’s surely cost competitive to just hire a human to do all those things.

The price needs to fall drastically below new-car territory before it’s competitive with manual human labour.

> ”it’s surely cost competitive to just hire a human to do all those things.”

The cost of labour varies hugely in different parts of the world. The cost of hiring someone in Switzerland is on the order of 100X more expensive compared to Bangladesh, for example.

With many countries currently in an anti-immigration political mindset and with birth rates declining globally, labour costs are likely to continue to increase in the future.

But once a technology like general-purpose humanoid robotics exists, it’s costs are only likely to decrease over time.

With joblessness on the rise and no signs of a reversal, there will always be people desperate for work, even at minimum wages. And there will be those so desperate, that they will willing to be exploited by agencies that will underpay them, for instance to get visa sponsorships. Humanoid robots won't be able to compete with that, at least, not in the near future.

Even in Switzerland, unemployment is on the rise - it's up by a massive 12.2% compared to last year[1].

The only way I see humanoid robots becoming a threat is a company with deep pockets mass manufactures them and subsidises them heavily that they can compete with desperate humans.

However, I doubt an actual competent robot could ever be that cheap in the near future. I mean, I still haven't come across a Roomba-style robot that's actually smart enough to detect which obstacles it can go over, or have a small robotic arm or something that can move light things that's in the way. Like say there's a sock on the floor, it should be able to simply move it out of the way and continue vacuuming; or say there's a wire, it should be able to determine whether or not it's safe to go over the wire instead of going around it. So until I see some real advancements in roombas, I remain skeptical about humanoid robots. And when we do get a humanoid robot that's clever enough to make sense of all the chaos in common households - and take care of it intelligently - you can bet that it won't come cheap.

[1] https://swisscareer.substack.com/p/the-myth-of-low-unemploym...

> it’s costs are only likely to decrease over time

What is this based on? We're well past a 50-ish year deflationary period in the cost of major appliances (refrigerators, washing machines, etc). We're pretty clearly at or near the end of the deflationary era for computers and computation. Automotive... speaks for itself. We're still there for televisions, surprisingly; but it looks like these technologies tend to have a handful of decades of rapid cost decrease, followed by a never-ending cost increase over time as the manufacturers consolidate and claim an ever-increasing margin.

The computer of 10 years ago is still a lot cheaper than a modern model. Deflation stops basically only if hardware advancement stops.

EVs are a great example: they keep getting cheaper for what they provide, even if the price stays the same. 200 miles of range 5 years ago is now 400+ miles of range today. Compared to ICEs, where advancement has stalled for the last 20 years, which seem like a worse deal every year.

> Compared to ICEs, where advancement has stalled for the last 20 years,

It hasn't really stalled: VVT, VCR, Cylinder deactivation that works properly, and start-stop becoming commonplace are all meaningful improvements (though smaller than the ones seen in EVs over the same time frame, which makes sense given the relative maturity).

ICE advancements haven’t materially affected car performance like EV advancements have. Start-stop is considered an annoyance to most car owners, for example, not a feature.

It is mostly because ICE tech is mature and has no real place to go for improvements beyond incremental refinement. EVs can ride the wave of battery tech advancements for another decade or two.

Crazy take: this only appears to be the case because industry incumbents tend to start applying mechanisms of gambling addiction to products; providing gratifications right on the lower threshold of expectations promotes dependency. New entrants to the car industry have no reason to adopt such a strategy, and so EV owners always look happier than ICE owners, but not crucially so, mysteriously trapping ICE owners in their thing despite EVs appearing massively better by joy as subjective proxy measurement for actual progress.

Even crazier take: Japanese companies always do this. Like knocking out features in an alternating fashion, so that you never get features A and B together, and such.

Honestly I think the biggest advancements in ICE cars recently have been the development and maturation of hybrid cars. Imagine telling someone in 2006 that your minivan got 36 miles per gallon!

36 mpg is 6.5 l/100km. That's only a smidge better than the most efficient Renault Espace from 1999: https://www.auto-abc.eu/Renault-Espace/g988-1997

You're comparing diesel to gasoline

Those are all tech that almost everyone owns. Makes sense that mass production would reduce costs and then the cost reductions would go towards zero. For new technology that hasn't been mass produced, it's a completely different story.

But that's my point. If you're basing your society shape around adopting a technology based on it continually decreasing in price, but you only get a few decades of that behavior before saturation and then you're at the mercy of the consolidated winners... generally adoptions like this aren't reversible at the societal level. You're locking in a long-term structural change based on a short-term pricing trend.

Still doesn't make sense. You were criticizing: "But once a technology like general-purpose humanoid robotics exists, it’s costs are only likely to decrease over time."

But the point was currently general-purpose humanoid robots is not affordable by the average person like say cars or washing machines currently are, but it will be because the costs will decrease. There was no argument that the price will forever keep going down, just like the price of cars or washing machines are not expected to constantly go down.

> costs are only likely to decrease over time

>> There was no argument that the price will forever keep going down

It's very hard for me to read the first quote as anything other than a continuing decrease in expected value of cost as a function of time. This directly contradicts the second quote.

Ok, this is definitely a reading comprehension issue that you have that you should fix somehow.

That’s only if you look at the final price without understanding the makeup of that price.

The components of computation have been getting cheaper every year… it hasn’t lately because the demand for memory suddenly massively started outstripping supply.

Just like RAM and disk prices have been continually decreasing for as long as people remember.

Humans are messy to deal with. Say you're rich enough to afford a personal chef. Unless you're an inhuman monster, their problems become your problems as well. So if your chef is out because their mother is sick and needs someone to take care of her, you pay for a nurse for your chef's mom, so that you have your chef. A robot servant is still gonna need maintenance, sure, but it's a bit easier to be callous to a robot than a person.

I think that’s a feature. Not a bug.

But for now, humanoids are still relatively incompetent. So if you're rich enough and you want things done like cleaning and cooking, it's more convenient to hire a human and give them specific instructions instead of buying a robot that can currently do only half the things they want to get done. Or they just want a nice gadget.

>but it's a bit easier to be callous to a robot than a person.

George Jetson always dreamed of beating a robot chef/maid.

A new car is what, $25-50K? Thats a one time cost.

How much do you think youd need to pay a maid every year to do your cooking, cleaning, laundry, dirty dishes etc? Coming once a day for 2 hours would be very expensive and still wouldnt be comparable to a robot that you own and is constantly deployed.

And at that point you’re probably comparing owning a robot to renting one.

Most people buying a new car don't pay them in one payment, they either lease or credit.

A car has a number of periodic maintenance to do + in most country some tax and some yearly technical inspection.

This is non counting the energy required by that car, electricity or gasoline

With the way subscription models and terms of use have been evolving over time, even if you buy a futuristic robotic assistant outright, you won’t really “own” it.

"A new car is what, $25-50K?"

Apparently $50K is closer to average in the U.S.

Price aside, the more important factor is that we don’t have the repair infrastructure to make something like this worthwhile yet. For something as critical as a car, we have workshops, spare parts supply chains, and the skilled technicians to do the repairs. Conventional robots require a similar skill set, but you still won’t be able to rely on a local repair for something people would expect to be dependable, like aged care or home assistance.

> it’s surely cost competitive to just hire a human to do all those things.

Depends on the service life/performance/etc.

As a simple benchmark, I will propose 'Mowing the lawn with a push mower'. Let's wave hands and assume there is a setup on a truck where the mower can be parked and then lifted in.

If you're paying the people doing that lawn-mowing federal minimum wage, at 40 hours a week it's 15K/year.

After 3 years that's 45K, or a little under the current US median price of a new car.

IOW, if the robot costs 45-50K, but can make it through 4 years without expensive maintenance you are still 'saving costs'.

There's hand-waving on both sides of my equation; At least where I live even pushing a lawnmower gets you a bit more than minimum wage (although it is more seasonal,) and also I have no clue if when we say 'new car territory' we are talking median or an 80K EV.

If you're paying cash/under the table, then maybe. But even then a twice a week household cleaning hire is going to cost upwards of $1500/mo unless you're being particularly exploitative. If you're not under the table, you're paying payroll taxes, probably paying for a payroll service, etc. so you're talking $2000+. At best you can maybe stay under $20k a year.

When you really look at the economics of it, a robot that never gets sick/doesn't require payroll/etc. makes a lot more sense.

This comment seems insane to me. Like at $50 an hour thats 30 hours a month, or 8 hoursish a week. How dirty or huge is the house? And $50 an hour is way over what most hourly degree edgucated workers earn so definitely not exploitation.

I pay $40/h for 4 hours once a month. If the wife had her way, it would be weekly.

But presumably would not be for four hours a week, since they would do more maintenance-like cleaning. Once a month, yeah, it’s going to take that long.

Why wouldn’t it take that long? When you hire people to do housekeeping they don’t just come for the exact amount of time you need them on an infrequent, case by case basis. You generally have to agree to a standard number of hours and days or a flat rate per day. They’re trying to make a living. If you can’t give them consistency and a meaningful amount of hours, they won’t work for you (or only will temporarily). They will find someone else who can provide more stability.

All that being said, what do you think this will cost? Several of you are scoffing at my numbers but I am very curious what you think they will leave their house for on a daily basis and how many days a month they would be coming by. Plus all the taxes and such that come with doing it legally.

If we’re trying to compare this to some sort of robot that does all your chores, then we have to at least start at 3 days a week. I’d compromise at 2 I guess but 8 days a month doesn’t seem like a fair comparison to a full-time robot chore handler.

If you clean a house once a month, it will take longer than doing the same thing once a week. You will end up spending more hours total, but it won't be four hours per visit, more like one or two. You spread the tasks out. Clean kitchen and den one visit. Clean bedrooms one visit. Clean bathrooms one visit. Or whatever.

The total cleaning time (and thus price) goes up, but it's not 4x what you're getting now, which a whole-house cleaning once a month would be.

Once a month and it’s a different task each time for an hour or two? No one does that - a monthly or bi-monthly cleaning service is $200-$400 and they generally do the whole place in one fell swoop and they’re definitely not going to agree to cutting their pay since you don’t want your full house serviced after taking the time to prep, come out, and maybe say no to another job. I don’t know if you have freelanced before, but this is how you think when you engage in contract work. “Time is money” applies 10x compared to a salaried job. All of this is to say 3-4hrs is very typical. You seem to be coming up with the single thinnest definition of what constitutes hiring for housework and it just doesn’t reflect 99% of how these hires go.

Also, we are comparing to a robot that does all your chores any day. Once a month is not a fair comparison to begin with. This is carrots to oranges.

Let’s be super generous here. $100/day twice a week for a basic pass at your house. Dishes, laundry, some wipe downs, put things way, standard chores. $800/mo, $9600/yr. Plus payroll service/taxes/etc. Hell do $100 once a week, which you’d be lucky to get. This is a $5-6k a year investment for the bare minimum and you’re still doing 80% of the chores annually. If you’ve got kids this person is barely making a dent.

Hiring people is expensive. If it isn’t, you’re not treating them with basic dignity.

Somehow wires got crossed here. I'm saying that once-a-week cleaning won't be 4 hours each time.

It is expensive, just not four times as expensive as a once-a-month job.

What sort of expenditure are you envisioning on a monthly basis?

> But even then a twice a week household cleaning hire is going to cost upwards of $1500/mo unless you're being particularly exploitative.

Sorry, what? Unless you're doing a deep clean of your house twice a week or you live in a particularly HCOL area, those numbers don't add up. You shouldn't be spending more than $1k/month on household chores, and even that seems high.

Source: A client of ours runs a "personal help" service (mostly focused on household tasks like laundry, tidying, organizing, etc as opposed to deep cleaning) so I have a lot of data on this. And they're a relatively premium service compared to some of the cheap labor you can actually buy. But they also don't operate in SF or NYC, so maybe prices are drastically different there.

$1k/month in an HCOL like here in Seattle doesn't give you much:

> In Seattle, hiring a house cleaner typically costs $150 to $500+ per visit, with most recurring standard cleanings for an average-sized home landing between $180 and $300. If you pay by the hour, rates generally range from $45 to $65 per hour for self-employed independent cleaners and $75 to $125 per hour for professional cleaning companies.

Even in high cost of living coastal Southern California these numbers are insane unless you have a $10mil house with 10,000sqft

Then tell me what you think it would cost. Hourly pay, how many hours they’d work in a day, and how many days a month you would expect them. What would be the monthly cost for the kind of help you are envisioning?

Not really? I pay $40/h, it takes them 4 hours to clean a 4 bed/2 bath. I do it once a month, but they’re talking twice a week.

Yeah I only see robomaids as an affordable option for someone that needs help with absolutely everything. These things are built out of commodity parts. Maybe you can make a robomaid a little cheaper if you build a lot of them to offset the upfront costs but not by much. Anytime the robomaid isn't working, it's just decreases the value of having one versus how much you paid for it. So the point would be to put it to work as much as possible such as for an elderly person that's unable to do anything for themselves.

Who spends new-car money to clean their homes? Maybe ultra high net worth individuals? I know people with 8 figures net worth who spend a fraction of that money for cleaning their homes.

Given the current context the presumption would be you have somebody coming 2 to 3 days a week. Are you telling me somebody’s going to come work for you for less than $100 a day 2-3 days a week? Why would they even take the gig?

My point is even 8 figures net worth individuals do not pay for someone to clean their house 2-3 times a week.

We set a bar for income brackets here as well as decided this is just for individuals?

None...

maintenance cost of machines is largely driven by human labor cost

It's robots all the way down.

unless you need rare earths (or any short supply / scarse resource)

extraction of which is driven by.......?

Around $80 - $100 an hour here in Sydney.

But a robot is much better than a human for these tasks. I feel uncomfortable paying a person to come to my house every day and clean. But a soulless robot, I don’t mind. Same for an in house private chef. Etc etc.

Also, like 10 neighbors could potentially share a robot and have it just go house to house every day.

See what you do is make it tele-operated. That way there is no stranger physically in your house, but technologically it's a lot easier to implement than some grand challenge-level laundry folding AI.

> Not something I would pay for to use outside of work

You wouldn't but apparently your employer would.

I don't disagree with you on robotics, though. For an empire like softbank, not buying an "insurance against the rise of robotics" also seems like a mistake to me.

That being said, they may expect robotics to rise through self-driving cars (hence their investment in Wayve).

> But a household robot that could clean, wash and fold the laundry, do the dishes, maybe even be a chauffeur... that would be huge.

Except they can't. I get it, merging advanced AI with robotics has made huge leaps in the past few years, but building a truly autonomous laundry robot is an incredibly difficult problem that still feels many years away. And I've seen all the "folding robots" over the past few years, and they are still miles away from being useful in your average home (they only "fold" if pieces are handed to them one-by-one, or the more advanced ones that can pick out clothes from a pile look like they were folded by a 3 year old).

Also, consider that all-in-one washer/dryer combos have existed for a while, but they are still a teeny percentage of washer sales because they're expensive and require more maintenance. There is a surprisingly low threshold on what people are willing to pay for labor saving devices.

> There is a surprisingly low threshold on what people are willing to pay for labor saving devices.

A lot of people gladly pay humans to launder their clothes, so the market is there. But the current iteration of wash/dry combo machines doesn’t solve the main issue. I don’t mind transferring my clothes from the washer to the dryer, because that takes 60 seconds. It’s folding and putting away the clothes that makes laundry a chore.

Getting to a cheap household helper robot requires building the expensive one. I don’t think Boston dynamics believes any reasonable consumer would buy their atlas robot, but by building it and scaling its production in industrial use cases they will learn things that make building a cheaper one much easier. And they will have built some factories to mass produce them. It’s not something that will be in everyone’s home next year, but sooner or later the robot hardware and robot intelligence will both be cheap enough to be accessible by average people (at least in the developed world)

Well, that assumes that if you just keep throwing more data and compute at large language models you'll end up with something akin to AGI to control those robots. Which is far from guaranteed.

LLMs aren't the specific architecture you'd use, but it very much looks like a tractable engineering problem to go from a university research lab project that can manage to fold clothes as a demo that needs to be played at 8x speed, to a sellable consumer product. The timeline is gonna be off, so no one knows if it's gonna take 3 years or 30, but it's not going to take an unknown breakthrough in materials science and physics the same way that nuclear fusion looks like it will require.

LLMs already solved the "System 2" part of this, to borrow from Kahneman, it's the "System 1" part that's lagging behind here. Current Claude/Gemini/ChatGPT is more than enough to tell a robot what chores to do, what to do with a thing, how, where to put it, etc. but what's still missing is the ability to reliably translate those goals to movements of a robot in diverse and tight environment that is a typical house or apartment, with any kind of reliability and safety.

No, you're assuming that you need AGI to control a robot, when LLMs have already shown you don't need anything close to hold a conversation.

So why do you suddenly think you need it for controlling a body when animals do it with far less?

But an animal has way more intelligence than an LLM? And unlike a chatbot, I can't delude myself into thinking it has done my laundry or fed my dog just because it tossed some clothes halfway into the washer without soap or spilled dog food on the floor. Unlike code, you can't just simply edit real life mistakes after the fact and call it good. My carpet can't be magically unstained, my glasses unbroken, or my dog or child unsmashed by the falling robot that had a bug fly in to a sensor at the wrong moment.

Is there any sign beyond flashy demos that humanoid robots will be functionally feasible though (before we even get to economically feasible)?

I know there's tons of activity on humanoid teleoperation data collection, and motion model training, but it hasn't seemed to bear out much of anything.

Like.... AI would be great if I could put it into a magical semi-corporeal familiar but I'm just not seeing a path to those either.

> ”maybe even be a chauffeur”

Cars will be able to safely drive themselves autonomously well before there is a humanoid robot capable of safely driving non-autonomous cars.

There are machines that already do those things. And if you’re rich enough to afford and maintain these humanoid robots, you would probably just hire staff already.

The only way I could see these AI robots take off is if on top of all those things, it could also perform sexual favors and develop personalities for people to bond with. Robosexuals would buy these primarily for those features and then household duties as nice plausible deniability.

> But a household robot that could clean, wash and fold the laundry […]

Agreed. Boston Dynamics’ focus has always been more on industrial uses, though, and while they’re getting to the point where eg. Atlas might be useful in a factory environment, they’re still a ‘premium’ supplier.

Seeing some of the stuff coming from Unitree and other competitors, SoftBank might be wondering if BD can stay in the game.

I would be more worried about the Chinese owning this market (like they did with robovacs) and not leaving much for Korea/USA outside of the defense market. We are still 5 years away from a general purpose functional household robot, but the rate of advances, even if they slow down substantially, will get us there.

The west can compete but lack the will to do so see the consumer drone market, note that attitude had better change or else the rest of the century is going to be hell.

Oh yes. We definitely have anti-intellectualism and a defeatist attitude to get over, or China will be the sole preeminent power very very soon

We're probably still decades away from that kind of robot. And I firmly believe the domestic robot revolution will not involve any humanoids. It'll be a few purpose-built machines that resemble nothing organic.

SoftBank has a history of making investment decisions that are the absolute opposite of good. I'm always bewildered as to how they still exist since they make nothing but absolute blunders.

They're not making household robots. They have a factory robot that only needs to operate in a controlled environment and boost profitability.

All of finance is trading money for time. $1 million today vs $100,000 for the next ten years. Softbank needs the money today vs later.

I think they would sell like hot cakes if they had attachments to also do mrs and mr after the chores ;-)

Don’t you think SoftBank has more insight and a ground level view on their own investment though?

SoftBank at times seems to be clueless with some of their investments, Arm Holdings was their best however...

Ok but that is pure fantasy.

Now how about a household robot that does all those things and is controlled by an Elon Musk company or by some other completely benevolent techno oligarch?

Reply on news.ycombinator.com