Yes, but not via robotics any time soon. The first wave of AI job replacement will come from letting the computers make the decisions while the meatpuppets (personified by whoever is lowest-costing at the time) turn the knobs. If you have the plumbing expert system ready to go why would you wait for the actuators to catch up? Slap AR glasses on 5000 kids/immigrants, get a couple of actual plumbers to go around town checking the work and let the AI dictate which tool they should use where and how. GPS was the zero to one, now for everything else.
Yeah, the real opportunity here is turning $18/hr unskilled labor into $28/hr semi-skilled labor.
Instead of someone with knowledge fixing a leak under your sink, it'll be some kid who has never turned a wrench before, following some LLM instructions.
We're slowly turning our whole workforce into the worst call center you've ever called -- a bunch of idiots reading from a script with no agency or skills of their own.
The opportunity in this case being paying the wrench turner $18/hr while charging $28/hr for this "system". Very forward thinking, humane, progrssive even. Heck, I bet you could even lobby for $7.50/hr depending on where you are.
Doing unlicensed plumbing or electrical work is a crime where I live. Not that a law necessarily stops everyone (that problem already exists today), but I genuinely have a hard time seeing how companies get states to relax these kinds of laws because they exist for good reason. It doesn't surprise me to see companies try to make inroads here, but it's an uphill battle I think.
When I was a kid, a landlord in my community decided to "fix" a broken heating unit at one of his properties himself to save money, and wound up killing all of his tenants due to CO poisoning. Those are the kinds of situations that we are risking by enabling DIY contract work.
Indeed, the question not whether its illegal, but who has more political power to maintain/change/ignore the law. Plumbers/electricians or companies that want to 'disrupt' things with low-wage AI-assisted labor. History tells us that any solution with lower costs will win that battle no matter how shitty they are.
True, but there are some key differences between this and Uber.
One is that an Uber rider is never put on a list of "can't ever ride on other services again", whereas if you try to sell a house that has unlicensed plumbing or electrical work done, it will show up on inspection and any smart buyer will walk (or run) away. Tons of real estate deals fall through because the seller did shoddy work like this and it will demonstrably reduce the value of your property because it's only worth what people are willing to pay.
This is also true for when you want an actual contractor to come work on your house. If a licensed contractor sees that you're trying to fix unlicensed work, they too will run away because they don't want to inherit the liability. I think the very red-tape nature of contracting work will be one of the things that protects it from this idiocy.
You jest, but if you think about it, it's not too hard to see Idiocracy as a movie about AI [1]. What do you think humanity is going to be like after thousands of years of no selective pressure on intelligence and no practical value to knowledge and education?
Idiocracy is the coming AI Utopia: your great-great-great grandchildren spending their days yukking it up "Ow, My Balls!" and having even dumber kids.
[1] IIRC, the backstory of the movie has the automation developed after humanity started getting stupid, to keep things from falling apart, but it seems more realistic for it to happen the other way around.
This made me remember a very old book posing that if alien intelligence is out there in the Universe, artificial intelligence should outnumber biological intelligence, as biological civilizations would tend to be short-lived.
Imagine if our benevolent AGI made Earth an utopia where everyone can achieve their full potential if they wish, then having first contact with another AGI. The alien AGI would ask if our AGI keeps these dumb humans as mementos, not realizing those humans as smart as they can possibly be.
Besides the AR glasses, this is roughly what already happens when PE rolls up a trade or clinic. One (licensed) expert oversees a bunch of less-qualified staff following a script. Just optimize the revenue and let schmucks pay expert prices.
This is already well underway with humans on both sides - especially in appliance repair (which is different but a similar technical/commercial model).
I have personally experienced several warranty calls for major appliances under warranty. The "tech" who shows up knows less about the appliance than I do, and the first thing they do is get on the phone with their own support call, who attempts to figure out what the issue is.
It's maddening for the consumer (me) because it (again, in my few experiences) doesn't produce a quality resolution quickly.
An actual appliance tech who still does service calls is like a unicorn for me - once I find one the are getting all my calls. My current guy for washers/driers tells me he is the remote tech for a few major brands, and spends a lot of his time each day on the phone with the onsite meat puppets.
Interestingly, that's equivalent to the not ready for primetime play replacing junior engineers with coding agents. But as someone with decades of product development and support under my belt, so many unintended consequences here and there.
You don't pay the plumber to replace the widget. You pay the plumber to look at the problem and know what widget to replace and the correct set of moves to do so without flooding the house. Good luck!
Cue the "Obviously, with a master plumber guiding them, they could do this!" non-counterpoint counterpoint because you've just destroyed the cost savings of not sending that master plumber to the site and if you try to make the master plumber multi-task, there goes the cognitive load burnout event horizon just like... Stay with me... REPLACING JUNIOR ENGINEERS WITH CODING AGENTS!
Kill me now. The Robot Apocalypse has already segued from tragedy to farce and the AI intelligentsia need to read more Adrian Tchaikovsky and less Eliezer Yudkowsky. I eagerly anticipate the downvotes once again for loving the science and technology of AI yet despising the breathless ridiculous hype on both sides of the AI wars. This is an alien mindset, I get it, but do try to convince me the error of my ways by the only way you know how to do.
I mean, if intelligence capabilities don’t get meaningfully better than today, sure.
Every discussion I see here is working under “when they are another order or two of magnitude better than today.”
No one is saying masters are getting replaced today. But “you pay the x to look at the problem and know the y and know z” is literally just intelligence over complicated tasks. It gets replaced if these things get smarter. And there is not really a good reason to think they won’t besides “I don’t want them to” or “my identity is tied to them not getting smarter”.
If you've ever watched a world class prodigy try to do some basic home repairs, and I have, the comedy and the tragedy come intermingled. Which is to say a pure intellect cannot bring the robot apocalypse without grounded agents in the real world.
But I welcome the attempt to build a domestic home repair robot as it's a valid attempt at getting AI better at something it's godawful at currently.
Everybody thrives best when luck is on their side :)
Even the pros flood a house now & then. You have to figure they have insurance just in case.
Once the dust settles (or the water drains) an interesting indicator might be whether there is any difference in insurance rates between human and robot plumbers.
This is soooo optimistic of them. Do they really think they can just swap a dude who's been fixing all kind of stuff that can happen in an installation with robots and AI?
"So, what do you think it is?"
"i don't know dude, I just have to do as the machine tells me, what's a pipe wrench?"
Is anyone willing to pay for a person without knowledge asking an AI or a robot to mess with their AC or pipes? At that point I'd rather do it myself with AI, at least I will care and be responsible about the results.
Of course they don't think it will work. Just like most AI startups, the end goal is investor money. All they have to do is make some wealthy people believe that they'll capture the $19B HVAC repair and $210B plumbing market and they're set. Like Elon with the flying car, it's another grift.
no theyre not. the founder has clearly never worked a day in maintenance-trades and doesn't understand the problem space at all. new construction residential plumbing is so cheap that there is no edge to having a robot do it, and emergency plumbing triage in finished homes is a game discovering exactly how wrong the usually non-existent schematics are.
Yeah, the number of times I have called a plumber over it's 100% been a troubleshooting operation requiring weird positioning, moving objects, turning things on and off. Maybe an LLM could do some of the thinking involved but there's basically no profit in replacing these people with machines, just not enough money in it.
Add required cooperation with electrician and potentially some more... yeah SV startup pipe dream.
I have proper professional plumber now doing repairs for a month in our apartment. He is part of 3 member team, the other is electrician with 40 years of experience and another tylist? but each can do each other's trade on cca comparable level, just due to personal connection I have with them).
For repair work, its always discovery, always more complex than anticipated, mostly non-standard, and the guy is creative, he has to be. And thats 20 year old apartment in heavily regulated Switzerland, nothing old or abscure.
Yep, and when you get beyond the basics its not just knowing what to do, but being certified to do it right so the house doesn't burn down and or flood.
It's incredibly hard to build a robot that washes dishes in a kitchen sink, let alone climbs up a rickety ladder into a garage attic and unblocks the HVAC overflow drain.
This will never be worth the risk. Every plumbing job is a unique challenge, you can't just spin up 10000 agents to brute force your way to a correctly cut pipe, one mistake and you've destroyed a house or office.
I tried to hire a general contractor or plumber recently. Called around, waited weeks for them to get back to me.
I ended up figuring out how to do the work myself.
What I gather is happening is that the K shaped economy is having wealthier people who benefit from the current tech boom buying and renovating houses at such a rate that skilled trades people are not bothering to chase smaller jobs.
That said, I am going to need some pretty good evidence a robot can be trusted to do the work. Working on houses in any form results in finding weird and unpredictable work from DIYers and prior generations doing things in radically different ways. My sink for example did not fit exactly the space the old cast iron sink sat in. What is the right decision: knock the spacers off the new stainless steel sink, or cut pieces of the countertop to make it fit? The next problem I had was the even once it fit, the faucet did not have clearance for the bolt which held it in place - requiring a trip to Home Depot to consult with their experts on how to handle it (turns out: a PVC pipe spacer cut to fit).
That is a lot of improvisation and dexterity for a simple sink replacement. It is going to be a while I think.
The interesting thing to me about this article is how the company tries to address AI/robot anxiety:
"Home-service technicians perform millions of highly technical tasks every day. We are creating a way for homeowners and service companies to benefit financially from the value of that expertise.
"We want to build the infrastructure that connects that knowledge with the companies building the next generation of intelligent machines while making sure the people creating the value participate in the economics."
While these technicians may be getting a bonus on top of what they usually charge, it's not like they're getting royalties or stock options for training their replacements. If you're a technician working or contracting for a service company that agrees to this, you might not even see any of that extra money.
It seems kind of silly with the current state of robotics, but in a way it's a kind of prisoner's dilemma: short-term gain of being paid to train the robots vs. long-term loss of being automated out of a job. Licensing is a barrier, but we let them drive cars in some cities.
If AI comes for the trades it seems like it will not be via robots but instead by leveling up all of the DIYers and lower skill handyman types allowing them to troubleshoot/build/fix more than they were capable of before.
Trades will be under attack on both fronts. The DIY kits will be used to train the robots. Don’t under estimate how motivated these companies are to embody AI and replace human labor.
If your insurance denies coverage when your DIY repair causes a loss, I think the advantage is overstated. Part of hiring a pro is the knowledge and tools, part of it is shifting the liability to someone insured and bonded.
I know enough building sciences to do most repairs or renovations myself. I still pay for a professional when I don’t want the liability risk (plumbing or electrical in a large multifamily property, for example). A few hundred dollars is always cheaper than my property insurance deductible, or the entire loss of coverage is denied (and property insurers are denying roughly half of annual claims).
I say never do you own plumbing or electrical and I stand by that. You’ll want to be able to show electrical work was inspected and approved by a master electrician god forbid an insurance event like a fire puts that into question.
AR glasses plus an LLM would be great for DIY maintenance, renovation tasks. Plumbing, electrical work, kitchen renovation, etc… often times it’s just knowing what to look for or check.
I have a cousin who has a certain fearlessness when it comes to home DIY projects. Unfortunately he could probably use a little more restraint, but on the whole it's an interesting quality that I noticed LLMs/coding agents help with quite a bit. They give me, an impostor syndrome type/overthinking type a certain confidence to move forward that really helps break into projects. I think this could probably be very effective in other domains like home improvement. And in case of my cousin, might also help avoid some missteps.
Yeah, this would be a much better version of the well-meaning but poorly produced youtube videos I typically use to fix things up around the house. Probably a lot of value there.
Assuming this would even work (unlikely), I suppose it's a good opportunity to take advantage of VC-subsidized plumbing repairs — if you're willing to take the risk.
Grok is just directly advertising the “let ais help you buy and run service business” foolishness as the next wave of home flipping. One good thing about interest rates being up is the capital to just hoover up a bunch local services will be harder to come by.
The last time I saw a news article about this it was talking about how many Airbnb kitchens in SF had been wrecked by teams running field tests for robots doing dishes. Excited to hear about the damage actual plumbing jobs would cause.
Nothing is inevitable. If this was inevitable, it's only under our current anti-human from of capitalism (it was always that way, AI just took the mask off).
> Who want's to pay an extra bundle for a plumber to come in and destroy a bunch of walls to fix a clogged pipe? No job is safe.
If the clog is so bad they can't snake it out or something, I don't think a robot is going to do a better job. Plumbers have to destroy walls because plumbing is often installed in inaccessible places.
I think the worst part of it is that we are forced to value ourselves based on our jobs. Of course, there's no easy to answer to what everyone would do if no one had to work again.
I don't think people would choose to fix toilets to survive if they didn't have to
> I don't think people would choose to fix toilets to survive if they didn't have to
If I hadn't gone into tech, I might have been a plumber. Plumbing is a deeply interesting topic, and it involves a lot of physics and fluid dynamics. Acting like a plumber is a desperate idiot with a wrench is not an accurate take, and is really rude. Also, Master Plumbers who operate their own businesses generally make >$150k/yr.
I worked with plumbers on a daily basis for a number of years. The masters were generally never actually plumbing for the most part. I would say 95% of them did not treat plumbing as an intellectual exercise. Of course, this is just an anecdote. I heard the "your shit is my bread and butter" hundreds of times from different plumbers thinking it was comedies final joke.
I feel like a lot of robotics companies are trying to ride the AI hype train right now, but the tech is soooooo far behind what we're seeing with knowledge work. The bridge from raw vision input to high level task definition to low level actuator movement with force feedback is super difficult. Unless we can get superintelligence to solve it for us, I don't see this happening for many decades.
AI is coming for a lot of physical labor, but it’s definitely not electricians, plumbers, roofers, barbers, etc. anytime soon. I genuinely believe that’s going to take at least a decade or two. Those jobs are way too precise and finicky for a robot to be able to do it. Cleaning a house, commercial restrooms, loading and unloading, delivery, landscaping, mail delivery, servers, chefs, old age care, truck drivers etc. will definitely be automated within the next five years though by robots.
No, I think within five years you’ll literally have a robot standing in front of a cooking range with the mise en place that another robot has prepared and will take orders and cook custom meals. Another robot will be picking up the dirty pots/pans/dishes, doing a rinse and putting them away into a dishwasher. For old age homes it will help people get out of bed. It will bring them their medication. It will prepare food for them. It will clean up the house that sort of thing. I think these robots at least for the short term will always be able to be teleoperated by a human in case of emergencies or mistakes, just like a Waymo, but those are going to be more and more infrequent as time goes on.
How quickly text based LLMs gained “intelligence” was shocking to all of us. What you need to understand about LLMs is they don’t care whether it’s text or a language or binary data. All they need to have is sufficient training data in order to be able to replicate human behavior. There are a lot of companies collecting training data for exactly these sort of real world activities now. When we get to a point where there’s enough training data where you can input a camera feed and the output is signal controls for the actuators on the robots, they’re going to start looking and acting like humans all of a sudden.
Thank you for your explanation. We have very different view of the future. Generating a correct piece of code can not be compared to taking care of an old, fragile individual.
Even if this robotic technology existed today (which it does not), it would take more than five years even to get it out there in the real world. If you think people in general are negative to AI, wait until they're told "AI will look after your elderly mother".
You’d be surprised how much progress has been made in the space in the last 3 to 6 months alone. It’s all very new but moving very fast.
As for what exists today, there is a ton of content online including very long live streams. Take a look at what figure robotics is doing with figure 3 (autonomous mail sorting, dishwashing, clothes folding, house cleaning streams), what unitree is doing with its speech to action LLM, as well as what Gemini robotics is doing. These are all on device autonomous physical LLMs that are in the very early stages and the stuff they can do is already incredible. You can look up Sunday robotics that essentially can fold 99% of clothes as shown on a 3 hour stream. There are a host of other very big players in the space, including Tesla robotics, Boston dynamics, and dozens of other US and Chinese robotic companies that are making a lot of progress very quickly.
> If you think people in general are negative to AI, wait until they're told "AI will look after your elderly mother".
People might be initially skeptical but if you tell them, it’ll happen for 1/10th of the price and people see this working for a few months they’ll quickly hop on the bandwagon. I don’t think you understand how expensive elder care is currently, people pay thousands of dollars per month for it.
the problem is no two houses are a like and you have to navigate, crawl spaces, basements, stairs, tiny utility closets where you can't even see the bolt you are tightening. there is no way this works in real everyday homes.
I’ve had two tract houses built, and I’ve found out that blueprints are mere suggestions on how to build the house. Sub-contractors do what they want. I even had a HVAC sub put chases in the middle of two closets to save money on ductwork; I was only able to get the super to move one since that was rework he had to pay for despite blueprints not being followed.
It's like if you took some of the challenges of waymo and made them even worse. I imagine it'll take a similar amount of time to get there (although waymo still hasn't rolled out anywhere 'unpredictable' e.g. Naples, Istanbul, Mumbai).
Self-driving is a great parallel. Whereas the tricky part of self-driving are the edge cases, maybe 5% of what the car needs to handle. What DIY plumbing I've done tells me is that plumbing is just about nothing but edge cases, with 5% being standard stuff that an LLM might be able to handle.
Yes, but not via robotics any time soon. The first wave of AI job replacement will come from letting the computers make the decisions while the meatpuppets (personified by whoever is lowest-costing at the time) turn the knobs. If you have the plumbing expert system ready to go why would you wait for the actuators to catch up? Slap AR glasses on 5000 kids/immigrants, get a couple of actual plumbers to go around town checking the work and let the AI dictate which tool they should use where and how. GPS was the zero to one, now for everything else.
Yeah, the real opportunity here is turning $18/hr unskilled labor into $28/hr semi-skilled labor.
Instead of someone with knowledge fixing a leak under your sink, it'll be some kid who has never turned a wrench before, following some LLM instructions.
We're slowly turning our whole workforce into the worst call center you've ever called -- a bunch of idiots reading from a script with no agency or skills of their own.
The opportunity in this case being paying the wrench turner $18/hr while charging $28/hr for this "system". Very forward thinking, humane, progrssive even. Heck, I bet you could even lobby for $7.50/hr depending on where you are.
Ain't that basically the plot of "Manna" by Marshall Brain?
Doing unlicensed plumbing or electrical work is a crime where I live. Not that a law necessarily stops everyone (that problem already exists today), but I genuinely have a hard time seeing how companies get states to relax these kinds of laws because they exist for good reason. It doesn't surprise me to see companies try to make inroads here, but it's an uphill battle I think.
When I was a kid, a landlord in my community decided to "fix" a broken heating unit at one of his properties himself to save money, and wound up killing all of his tenants due to CO poisoning. Those are the kinds of situations that we are risking by enabling DIY contract work.
There's usually exceptions for simple work -- changes to fixtures, removing clogs, fixing minor leaks, etc.
But of course, with enough VC money you can take the Uber route and 'creatively interpret' the law
Uber was once illegal everywhere. So were horseless carriages.
Indeed, the question not whether its illegal, but who has more political power to maintain/change/ignore the law. Plumbers/electricians or companies that want to 'disrupt' things with low-wage AI-assisted labor. History tells us that any solution with lower costs will win that battle no matter how shitty they are.
True, but there are some key differences between this and Uber.
One is that an Uber rider is never put on a list of "can't ever ride on other services again", whereas if you try to sell a house that has unlicensed plumbing or electrical work done, it will show up on inspection and any smart buyer will walk (or run) away. Tons of real estate deals fall through because the seller did shoddy work like this and it will demonstrably reduce the value of your property because it's only worth what people are willing to pay.
This is also true for when you want an actual contractor to come work on your house. If a licensed contractor sees that you're trying to fix unlicensed work, they too will run away because they don't want to inherit the liability. I think the very red-tape nature of contracting work will be one of the things that protects it from this idiocy.
Welcome to Costco. I love you.
> Welcome to Costco. I love you.
You jest, but if you think about it, it's not too hard to see Idiocracy as a movie about AI [1]. What do you think humanity is going to be like after thousands of years of no selective pressure on intelligence and no practical value to knowledge and education?
Idiocracy is the coming AI Utopia: your great-great-great grandchildren spending their days yukking it up "Ow, My Balls!" and having even dumber kids.
[1] IIRC, the backstory of the movie has the automation developed after humanity started getting stupid, to keep things from falling apart, but it seems more realistic for it to happen the other way around.
Well, humanity had started becoming stupid but there were enough smart folks to develop the automation. Then they went extinct.
Your floor is now clean.
This made me remember a very old book posing that if alien intelligence is out there in the Universe, artificial intelligence should outnumber biological intelligence, as biological civilizations would tend to be short-lived.
Imagine if our benevolent AGI made Earth an utopia where everyone can achieve their full potential if they wish, then having first contact with another AGI. The alien AGI would ask if our AGI keeps these dumb humans as mementos, not realizing those humans as smart as they can possibly be.
https://archive.org/details/intelligenceinun0000macg
Besides the AR glasses, this is roughly what already happens when PE rolls up a trade or clinic. One (licensed) expert oversees a bunch of less-qualified staff following a script. Just optimize the revenue and let schmucks pay expert prices.
This is already well underway with humans on both sides - especially in appliance repair (which is different but a similar technical/commercial model).
I have personally experienced several warranty calls for major appliances under warranty. The "tech" who shows up knows less about the appliance than I do, and the first thing they do is get on the phone with their own support call, who attempts to figure out what the issue is.
It's maddening for the consumer (me) because it (again, in my few experiences) doesn't produce a quality resolution quickly.
An actual appliance tech who still does service calls is like a unicorn for me - once I find one the are getting all my calls. My current guy for washers/driers tells me he is the remote tech for a few major brands, and spends a lot of his time each day on the phone with the onsite meat puppets.
Yes, or deliver me a kit of the exact tools and parts and AR glasses and I'll do it myself.
> the plumbing expert system
Expert system? What's this, the 80s? :)
But yes, I agree with the main sentiment. That's basically what Amazon already is (and Uber).
Interestingly, that's equivalent to the not ready for primetime play replacing junior engineers with coding agents. But as someone with decades of product development and support under my belt, so many unintended consequences here and there.
You don't pay the plumber to replace the widget. You pay the plumber to look at the problem and know what widget to replace and the correct set of moves to do so without flooding the house. Good luck!
Cue the "Obviously, with a master plumber guiding them, they could do this!" non-counterpoint counterpoint because you've just destroyed the cost savings of not sending that master plumber to the site and if you try to make the master plumber multi-task, there goes the cognitive load burnout event horizon just like... Stay with me... REPLACING JUNIOR ENGINEERS WITH CODING AGENTS!
Kill me now. The Robot Apocalypse has already segued from tragedy to farce and the AI intelligentsia need to read more Adrian Tchaikovsky and less Eliezer Yudkowsky. I eagerly anticipate the downvotes once again for loving the science and technology of AI yet despising the breathless ridiculous hype on both sides of the AI wars. This is an alien mindset, I get it, but do try to convince me the error of my ways by the only way you know how to do.
I mean, if intelligence capabilities don’t get meaningfully better than today, sure.
Every discussion I see here is working under “when they are another order or two of magnitude better than today.”
No one is saying masters are getting replaced today. But “you pay the x to look at the problem and know the y and know z” is literally just intelligence over complicated tasks. It gets replaced if these things get smarter. And there is not really a good reason to think they won’t besides “I don’t want them to” or “my identity is tied to them not getting smarter”.
If you've ever watched a world class prodigy try to do some basic home repairs, and I have, the comedy and the tragedy come intermingled. Which is to say a pure intellect cannot bring the robot apocalypse without grounded agents in the real world.
But I welcome the attempt to build a domestic home repair robot as it's a valid attempt at getting AI better at something it's godawful at currently.
>without flooding the house. Good luck!
Everybody thrives best when luck is on their side :)
Even the pros flood a house now & then. You have to figure they have insurance just in case.
Once the dust settles (or the water drains) an interesting indicator might be whether there is any difference in insurance rates between human and robot plumbers.
Indeed, any asset is better understood once it has been marked to market.
> meatpuppets (personified by whoever is lowest-costing at the time)
I prefer the Chernobyl name for them - biorobots
https://www.youtube.com/watch?v=ZL0rdf_3Ivs
This is soooo optimistic of them. Do they really think they can just swap a dude who's been fixing all kind of stuff that can happen in an installation with robots and AI?
"So, what do you think it is?"
"i don't know dude, I just have to do as the machine tells me, what's a pipe wrench?"
Is anyone willing to pay for a person without knowledge asking an AI or a robot to mess with their AC or pipes? At that point I'd rather do it myself with AI, at least I will care and be responsible about the results.
Of course they don't think it will work. Just like most AI startups, the end goal is investor money. All they have to do is make some wealthy people believe that they'll capture the $19B HVAC repair and $210B plumbing market and they're set. Like Elon with the flying car, it's another grift.
no theyre not. the founder has clearly never worked a day in maintenance-trades and doesn't understand the problem space at all. new construction residential plumbing is so cheap that there is no edge to having a robot do it, and emergency plumbing triage in finished homes is a game discovering exactly how wrong the usually non-existent schematics are.
Yeah, the number of times I have called a plumber over it's 100% been a troubleshooting operation requiring weird positioning, moving objects, turning things on and off. Maybe an LLM could do some of the thinking involved but there's basically no profit in replacing these people with machines, just not enough money in it.
Robobutler flickers shutters on its cameras. "I know plumbing. That would be $200/month."
Add required cooperation with electrician and potentially some more... yeah SV startup pipe dream.
I have proper professional plumber now doing repairs for a month in our apartment. He is part of 3 member team, the other is electrician with 40 years of experience and another tylist? but each can do each other's trade on cca comparable level, just due to personal connection I have with them).
For repair work, its always discovery, always more complex than anticipated, mostly non-standard, and the guy is creative, he has to be. And thats 20 year old apartment in heavily regulated Switzerland, nothing old or abscure.
Yep, and when you get beyond the basics its not just knowing what to do, but being certified to do it right so the house doesn't burn down and or flood.
It's incredibly hard to build a robot that washes dishes in a kitchen sink, let alone climbs up a rickety ladder into a garage attic and unblocks the HVAC overflow drain.
This will never be worth the risk. Every plumbing job is a unique challenge, you can't just spin up 10000 agents to brute force your way to a correctly cut pipe, one mistake and you've destroyed a house or office.
... or ten.
I tried to hire a general contractor or plumber recently. Called around, waited weeks for them to get back to me.
I ended up figuring out how to do the work myself.
What I gather is happening is that the K shaped economy is having wealthier people who benefit from the current tech boom buying and renovating houses at such a rate that skilled trades people are not bothering to chase smaller jobs.
That said, I am going to need some pretty good evidence a robot can be trusted to do the work. Working on houses in any form results in finding weird and unpredictable work from DIYers and prior generations doing things in radically different ways. My sink for example did not fit exactly the space the old cast iron sink sat in. What is the right decision: knock the spacers off the new stainless steel sink, or cut pieces of the countertop to make it fit? The next problem I had was the even once it fit, the faucet did not have clearance for the bolt which held it in place - requiring a trip to Home Depot to consult with their experts on how to handle it (turns out: a PVC pipe spacer cut to fit).
That is a lot of improvisation and dexterity for a simple sink replacement. It is going to be a while I think.
The interesting thing to me about this article is how the company tries to address AI/robot anxiety:
"Home-service technicians perform millions of highly technical tasks every day. We are creating a way for homeowners and service companies to benefit financially from the value of that expertise.
"We want to build the infrastructure that connects that knowledge with the companies building the next generation of intelligent machines while making sure the people creating the value participate in the economics."
While these technicians may be getting a bonus on top of what they usually charge, it's not like they're getting royalties or stock options for training their replacements. If you're a technician working or contracting for a service company that agrees to this, you might not even see any of that extra money.
It seems kind of silly with the current state of robotics, but in a way it's a kind of prisoner's dilemma: short-term gain of being paid to train the robots vs. long-term loss of being automated out of a job. Licensing is a barrier, but we let them drive cars in some cities.
If AI comes for the trades it seems like it will not be via robots but instead by leveling up all of the DIYers and lower skill handyman types allowing them to troubleshoot/build/fix more than they were capable of before.
Trades will be under attack on both fronts. The DIY kits will be used to train the robots. Don’t under estimate how motivated these companies are to embody AI and replace human labor.
If your insurance denies coverage when your DIY repair causes a loss, I think the advantage is overstated. Part of hiring a pro is the knowledge and tools, part of it is shifting the liability to someone insured and bonded.
I know enough building sciences to do most repairs or renovations myself. I still pay for a professional when I don’t want the liability risk (plumbing or electrical in a large multifamily property, for example). A few hundred dollars is always cheaper than my property insurance deductible, or the entire loss of coverage is denied (and property insurers are denying roughly half of annual claims).
I say never do you own plumbing or electrical and I stand by that. You’ll want to be able to show electrical work was inspected and approved by a master electrician god forbid an insurance event like a fire puts that into question.
The AI labs will pay for insurance as part of cost of doing business.
Who pays when the AI company has declared bankruptcy and problems were only realized after 5 years?
Until they do, it’s wild speculation. “Prove it.” Most companies cannot get AI insurance for AI use today before considering meat space liability.
AI Runs Into Its Next Big Obstacle: Insurance - https://news.ycombinator.com/item?id=50154288 - October 2026
AR glasses plus an LLM would be great for DIY maintenance, renovation tasks. Plumbing, electrical work, kitchen renovation, etc… often times it’s just knowing what to look for or check.
I have a cousin who has a certain fearlessness when it comes to home DIY projects. Unfortunately he could probably use a little more restraint, but on the whole it's an interesting quality that I noticed LLMs/coding agents help with quite a bit. They give me, an impostor syndrome type/overthinking type a certain confidence to move forward that really helps break into projects. I think this could probably be very effective in other domains like home improvement. And in case of my cousin, might also help avoid some missteps.
Yeah, this would be a much better version of the well-meaning but poorly produced youtube videos I typically use to fix things up around the house. Probably a lot of value there.
Doing electric work based on solely on LLM output sounds like a recipe for disaster.
Assuming this would even work (unlikely), I suppose it's a good opportunity to take advantage of VC-subsidized plumbing repairs — if you're willing to take the risk.
Grok is just directly advertising the “let ais help you buy and run service business” foolishness as the next wave of home flipping. One good thing about interest rates being up is the capital to just hoover up a bunch local services will be harder to come by.
The last time I saw a news article about this it was talking about how many Airbnb kitchens in SF had been wrecked by teams running field tests for robots doing dishes. Excited to hear about the damage actual plumbing jobs would cause.
It was inevitable. Who want's to pay an extra bundle for a plumber to come in and destroy a bunch of walls to fix a clogged pipe? No job is safe.
> It was inevitable.
Nothing is inevitable. If this was inevitable, it's only under our current anti-human from of capitalism (it was always that way, AI just took the mask off).
> Who want's to pay an extra bundle for a plumber to come in and destroy a bunch of walls to fix a clogged pipe? No job is safe.
If the clog is so bad they can't snake it out or something, I don't think a robot is going to do a better job. Plumbers have to destroy walls because plumbing is often installed in inaccessible places.
If you really want cheap plumbing, you could always move your pipes outside the walls (sort of like the ductwork in Brazil, https://www.youtube.com/watch?v=K9gO01pyv24).
I think the worst part of it is that we are forced to value ourselves based on our jobs. Of course, there's no easy to answer to what everyone would do if no one had to work again.
I don't think people would choose to fix toilets to survive if they didn't have to
> I don't think people would choose to fix toilets to survive if they didn't have to
If I hadn't gone into tech, I might have been a plumber. Plumbing is a deeply interesting topic, and it involves a lot of physics and fluid dynamics. Acting like a plumber is a desperate idiot with a wrench is not an accurate take, and is really rude. Also, Master Plumbers who operate their own businesses generally make >$150k/yr.
I worked with plumbers on a daily basis for a number of years. The masters were generally never actually plumbing for the most part. I would say 95% of them did not treat plumbing as an intellectual exercise. Of course, this is just an anecdote. I heard the "your shit is my bread and butter" hundreds of times from different plumbers thinking it was comedies final joke.
I would never hire an AI plumber. I want the guy who came here last time, did a good job, and pay him premium prices.
I feel like a lot of robotics companies are trying to ride the AI hype train right now, but the tech is soooooo far behind what we're seeing with knowledge work. The bridge from raw vision input to high level task definition to low level actuator movement with force feedback is super difficult. Unless we can get superintelligence to solve it for us, I don't see this happening for many decades.
Yep. Tesla's original value prop for the gigafactory was to have the machine build the machine. But that failed because wires are hard to install.
Good
Home repairs are too expensive
If I were in the trades, my larger concern than AR goggled meatpuppets would be:
1) The dramatic drop in demand for my services when most of the knowledge workers lose their jobs and thus incomes.
2) The dramatic increase in supply of people doing my job when all those knowledge workers attempt to flee to 'safe' jobs.
Crushing forces coming from both sides are likely to harm even the jobs that seem to have the lease AI exposure.
AI is coming for a lot of physical labor, but it’s definitely not electricians, plumbers, roofers, barbers, etc. anytime soon. I genuinely believe that’s going to take at least a decade or two. Those jobs are way too precise and finicky for a robot to be able to do it. Cleaning a house, commercial restrooms, loading and unloading, delivery, landscaping, mail delivery, servers, chefs, old age care, truck drivers etc. will definitely be automated within the next five years though by robots.
I want to understand. Do you seriously believe that in five years chefs, old age care, house cleaning etc will be "definitely" automated?
What do you mean by automated? No humans in the loop, or that the human doing the work is using some kind of AI tool to aid them?
No, I think within five years you’ll literally have a robot standing in front of a cooking range with the mise en place that another robot has prepared and will take orders and cook custom meals. Another robot will be picking up the dirty pots/pans/dishes, doing a rinse and putting them away into a dishwasher. For old age homes it will help people get out of bed. It will bring them their medication. It will prepare food for them. It will clean up the house that sort of thing. I think these robots at least for the short term will always be able to be teleoperated by a human in case of emergencies or mistakes, just like a Waymo, but those are going to be more and more infrequent as time goes on.
How quickly text based LLMs gained “intelligence” was shocking to all of us. What you need to understand about LLMs is they don’t care whether it’s text or a language or binary data. All they need to have is sufficient training data in order to be able to replicate human behavior. There are a lot of companies collecting training data for exactly these sort of real world activities now. When we get to a point where there’s enough training data where you can input a camera feed and the output is signal controls for the actuators on the robots, they’re going to start looking and acting like humans all of a sudden.
Thank you for your explanation. We have very different view of the future. Generating a correct piece of code can not be compared to taking care of an old, fragile individual.
Even if this robotic technology existed today (which it does not), it would take more than five years even to get it out there in the real world. If you think people in general are negative to AI, wait until they're told "AI will look after your elderly mother".
You’d be surprised how much progress has been made in the space in the last 3 to 6 months alone. It’s all very new but moving very fast.
As for what exists today, there is a ton of content online including very long live streams. Take a look at what figure robotics is doing with figure 3 (autonomous mail sorting, dishwashing, clothes folding, house cleaning streams), what unitree is doing with its speech to action LLM, as well as what Gemini robotics is doing. These are all on device autonomous physical LLMs that are in the very early stages and the stuff they can do is already incredible. You can look up Sunday robotics that essentially can fold 99% of clothes as shown on a 3 hour stream. There are a host of other very big players in the space, including Tesla robotics, Boston dynamics, and dozens of other US and Chinese robotic companies that are making a lot of progress very quickly.
> If you think people in general are negative to AI, wait until they're told "AI will look after your elderly mother".
People might be initially skeptical but if you tell them, it’ll happen for 1/10th of the price and people see this working for a few months they’ll quickly hop on the bandwagon. I don’t think you understand how expensive elder care is currently, people pay thousands of dollars per month for it.
the problem is no two houses are a like and you have to navigate, crawl spaces, basements, stairs, tiny utility closets where you can't even see the bolt you are tightening. there is no way this works in real everyday homes.
I’ve had two tract houses built, and I’ve found out that blueprints are mere suggestions on how to build the house. Sub-contractors do what they want. I even had a HVAC sub put chases in the middle of two closets to save money on ductwork; I was only able to get the super to move one since that was rework he had to pay for despite blueprints not being followed.
It's like if you took some of the challenges of waymo and made them even worse. I imagine it'll take a similar amount of time to get there (although waymo still hasn't rolled out anywhere 'unpredictable' e.g. Naples, Istanbul, Mumbai).
Self-driving is a great parallel. Whereas the tricky part of self-driving are the edge cases, maybe 5% of what the car needs to handle. What DIY plumbing I've done tells me is that plumbing is just about nothing but edge cases, with 5% being standard stuff that an LLM might be able to handle.
> AI lab offers free plumbing and HVAC repairs across three cities
Cool! VC capital at work for the common people.
> to train AI robots
LOL total bullshit.
Lol looks like no one is safe
[dead]
It's almost like they'll do literally anything to not pay you.
I'll believe this when I see it.