I was a very early customer of Roomba and loved them when they came out. I had pets at the time, and the machine would consistently fail in about 14 months. I finally figured out that I needed to buy them from Costco, so that I could get them replaced.
Rather than taking their lead and improving the product, they just sat there with the exact same product for like 10+ years. It was outrageous.
I guess Rodney Brooks got busy with other interests, and whomever ran things didn't realize that Tim Ferris is full of shit.
It was extremely frustrating to watch these assholes destroy the company right from the outset. All they needed to do, was to slowly walk forward and iterate with improvements.
The only surprise in this news is that it took SO LONG for them to dismantle the company.
I do not think it's appropriate for an organization holding this much deeply personal data can be sold to any foreign entity.
Rather than taking their lead and improving the product, they just sat there with the exact same product for like 10+ years. It was outrageous.
I think that this is actually the only viable strategy for a hardware product company in the current world.
As soon as your product is successful, it will be cloned by dozens of Chinese companies and dumped on the market everywhere. Any update you make from there on out will immediately be folded into all those products selling for 10% what you do. In a couple years, they'll all be better than yours, and still way cheaper.
So you have to do the Roomba thing or the GoPro thing, where you iterate behind the scenes until your thing is amazing, release it with a big Hollywood launch, get it turned into the noun and verb for your product category and the action that it does.
But then you have to do what those companies didn't do: Fire everybody and rake in as much cash as possible before the inevitable flood of clones drowns you.
I have a few really good hardware ideas, but I don't believe I could ever market them fast enough and far enough to make it worth spending the R&D to make them happen.
> I have a few really good hardware ideas, but I don't believe I could ever market them fast enough and far enough to make it worth spending the R&D to make them happen.
Yeah, almost a decade ago I had a dream of creating a drone startup with some very specific tech that would have required several years of R&D to create. The end product would have been relatively cheap to manufacture, being basically a PCB with a large FPGA plus a bunch of relatively cheap sensors.
I actually got about 6 months into the project, and then realised that although it was a great project and if it worked well, I'd be able to make units for about 25% of a viable RRP and be able to recoup all my time doing R&D without an income with maybe 5k units in direct sales. And then it slowly dawned on me that if I could build it for 25% of a viable RRP, then the Chinese cloners could do it even cheaper, and all they'd have to do was reverse engineer the protection on the FPGA bitstream to clone it and clone a pretty simple PCB. At the time, the drone market was full of cloned components for a fraction of the price of the original price, or of open source projects sold for half the price of the official boards to support the project.
In such a situation, the only way to really survive is to innovate faster than the cloners can copy it, but that's kind of predicated on making a product that you know isn't what you want the final product to be from the start, so that you can drip feed the improvements into the market every time the previous version was cloned. That would also have the side effect of alienating the early adopters, as well as making new customers wonder how long it'd be before the new product was obsolete. Ultimately, I decided that realistically it wasn't viable to continue doing R&D for another couple of years, unsure if I'd actually be able to pay myself going forward.
Having done the hardware game, it's not so much the clones that get you, it's the VCs/shareholders.
You need a lot of money to make hardware, so you get vc money and eventually shareholder money. But if you're not selling new hardware all the time, the company isn't making money. So they dictate that you need to make new hardware, yearly.
Making new hardware yearly is enough of an undertaking that you no longer have time to iterate on the software that could enable new features. And often hardware iterations aren't going to change that much, it's hard to "invent" new hardware. It's better to make a hardware platform that enables new exciting features, and iterate on the software. But that isn't going to sell yearly.
So unless you have a software subscription model that people love, every hardware company tends to stagnate because they are too busy making hardware yearly to make "better" products.
You see this very clearly in cameras vs phones. The camera companies are still making cameras yearly but none of them incorporate the software features that have led phones to outpace them. A lot of phones with so so cameras take better pictures (to the average eye) than actual cameras because the software features enhance the photos.
I worked on firmware for such a "noun and verb" product that IPOd a decade ago, and lived the struggle realtime.
> If society no longer values these qualities, then we don't deserve better.
Isn't it more like "if society has time to think about and can afford those qualities"?
If most folks out there have limited finances (CoL-relative, of course) and are just scrapping by, they'll buy the cheapest thing out there that just does the job (vacuums) and tend to ignore any extra luxuries, even if those would be more economically advantageous long-term (repairs/maintenance part of the TCO). That's simply because of the focus - it's more on the account balance, due bills and next paycheck, than on the implications for a more distant future. Crazy volatility and all the global rollercoasters like pandemics, wars, and all the crazy politicians around the world doesn't help regular folks' sensible decision-making at all, of course. The more stressed one is, the less rational they act.
People don't buy cheap junk because they don't value quality. They buy it primarily because of affordability reasons, or because their focus is forced to be elsewhere.
>People don't buy cheap junk because they don't value quality. They buy it primarily because of affordability reasons, or because their focus is forced to be elsewhere.
The focus, thanks to years of advertising, is shifted towards features, new features sell, quality doesn’t, so to keep the price point and “innovate” manufacturers need to lower the quality knowing that a new version will replace the device soon, most consumers see this as normal so when a poorly designed and cheaply made thing for what there’s no replacement parts , no repair info, no software/ firmware fails is just an excuse to purchase the new shiny iteration with all the the bells and whistles (and AI!, copilot toaster!) wich is gonna last les than the previous one but now needs an “app” an activation and a subscription for the premium features .
This isn't unique to China, it's just the nature of modern manufacturing. The only reason China stands out is because we offshored our manu there, so it's where we see it happen.
I feel like people forget that the entire purpose of factories/ automation/ modern manufacturing was to divorce human skill from product worth (so that companies wouldn't have to pay workers based on skill). That also means that in the realm of physical goods, "moats" are not maintainable unless you have a manufacturing technique or technology that others don't. Since companies rarely create their own production line machinery, anyone else who can afford the same machines can produce the same products.
The actual "viable strategy for hardware companies" has to be about market penetration; make products that aren't on Amazon, for example, and Amazon can't be used to out-maneuver you. Firearms are a great example of where manufacturing capability does not equal competitiveness; China can absolutely produce any firearm that you can buy in the US, but they don't because other factors (mostly related to regulatory controls) created a moat for manufacturers. Vehicles are another good example. Good luck buying an Avatr car in the US.
But yes, if you plan to make a vacuum, which is just you iterating on what others have done as well, you should probably expect that people are going to trivially iterate on your variant too.
While those patents are not enforceable in China (unless equivalents were also filed in China -- unsure if they would be worth much) they would be when imported to the US. This is one of the reasons the ITC exists, and it played a prominent role during the smartphone patent wars. So at least the US market would be protected from knock-offs.
The smartphone wars were fought among tech giants, not capital intensive hardware startups. The problem with patents is that you need to already be financially successful enough to file them, able to pay to protect them in court, and can float your company's operating costs long enough to see them enforced and rewarded, which may take years.
Yes and no -- filing patents is quite affordable (probably outdated info, but I recall average costs for drafting and filing was ~10K / patent, most of the costs being related to the drafting rather than filing.) Compared to all the other capital investments required for hardware startups, these costs are negligible.
But you're totally right that enforcing them is extremely expensive, slow and risky.
That said, Roomba isn't exactly a startup but wasn't a tech giant either, and did enforce their patents often.
And especially against imported infringing products, the ITC provides a much cheaper, faster mechanism to get protection via injunctions.
That's why the ITC is so relevant here: it is relatively quite speedy compared to regular patent trials, and have the power to issue injunctions against imports (which is partly why it was relied on a lot during the smartphone patent wars.) So you may not collect damages from Chinese companies, but you can completely block their infringing imports into the US and deny them US revenue.
Coasting on their patents is exactly why iRobot went bankrupt. If they had a proper incentive to continue innovating, they might be around today. Instead, the patent system incentivized them to erect a tollgate and snooze away in the booth next to it.
>> US companies can’t beat Chinese companies completely subsidized by their national government.
> Except our companies do just that, all the time. Who is the Chinese Intel? The Chinese Microsoft? The Chinese Boeing? The Chinese NVIDIA?
Where are the new ones?
Also Intel is not doing well, and the Chinese (after a fashion) Intel is TSMC, who also does NVIDIA's manufacturing.
> People forget that the US is still the #2 manufacturer in the world, and that's (apparently) without halfway trying.
So? That fact sounds like pablum. I think the real story of US manufacturing has been one of erosion of capabilities and long-term loss of strength. The US may still have a high ranking, but I'd bet: 1) much of that of that is low-volume and legacy, 2) second-place is still only 60% of what China does.
> People forget that the US is still the #2 manufacturer in the world
Manufacturer of what, exactly, though?
What do you export? What do you sell?
Food? Nope, illegal in most of the world.
Cars? Nope, uncompetitive in most of the world. "High end" American cars lack even basic features fitted to poverty-spec cars in the EU, like heated windscreens.
Computers? I'm typing this on a computer assembled in Scotland onto a Latvian-made chassis using a Chinese-made motherboard populated with Korean memory chips and an Israeli microprocessor.
What does the US actually make and sell, any more?
Zhaoxin makes X86 and countless make ARM and RiscV chips. SMIC being a foundry.
> The Chinese Microsoft?
Baidu, Tencent, Alibaba, ByteDance.
>The Chinese Boeing?
Comac makes passenger and Chengdu fighter jets.
>The Chinese NVIDIA?
Huawei makes AI GPUs.
>People forget that the US is still the #2 manufacturer in the world
Considering the US never had its industry blown up in any war and could reap the benefits of 150+ years worth of stability, higher education, skilled immigration, compounding wealth, and taking over the vacuum and brains of Europe's post-war industrial powers, that's not really something THAT impressive.
>and that's (apparently) without halfway trying.
If it isn't halfway trying, why does it feel the need to sanction or ban chinese competitors?
> If you're really the first, you should be able to get about a 20 year head start.
That's an opinion, and not one I agree with.
If you and your competitor are racing to develop a thing, whoever wins by a couple months shouldn't get a monopoly for decades.
Most of the time when things get patented, it's strictly worse for innovation in that space until the patents expire. 3d printing is a great example.
It's asinine to think you can outsource manufacturing of whatever object to some other company in another country, but that no one on the planet can make the same thing because "the idea is yours".
> Most of the time when things get patented, it's strictly worse for innovation in that space until the patents expire.
What happens at expiration is an important and intended feature of patents. They trade a legally guaranteed headstart against the requirement of publishing your methods for your competitors to learn from.
> I have a few really good hardware ideas, but I don't believe I could ever market them fast enough and far enough to make it worth spending the R&D to make them happen.
>I have a few really good hardware ideas, but I don't believe I could ever market them fast enough and far enough to make it worth spending the R&D to make them happen.
Then make a nice blog post, translate it to Chinese (hell, I'll pay a professional translator for you) and post it on the internet so that someone in Shenzen can try it.
Just post your ideas to crowdsource websites and wait for the aliexpress clone to appear, zero r&d costs, zero dev and manufacturing/qa! That said, Taobao and Ali are so full of bizarre products (transparent rubber domes to be able to type with 5cm long nail extensions), it will be a challenge to stand out
I wonder why nobody has tried to beat the Chinese companies at their own game. The whole schtick is: take a product that people like, vertically integrate and drive down costs. This is like the purest form of capitalism.
* built a lithium refinery
* produces its own battery cells
* makes its own motors and drivetrains
* makes its own car seats
* owns and operates a fast-charging network
* sells direct, bypassing dealerships
* offers insurance integrated with vehicle data
* develops its own autopilot AI
Great point, and to drive it home -- TSLA is the only competitive non-Chinese company in the EV space. You could make the argument that it's one of very few successful U.S. manufacturing company winning on purely technical/capitalist terms, considering the whole U.S.-Taiwan stranglehold on chip mfg
> You could make the argument that it's one of very few successful U.S. manufacturing company winning on purely technical/capitalist terms
Except it's not winning on that at all. It's "winning" because Chinese EV brands are barred from selling in the US. You can't buy an Avatr if you want. It's in fact protectionist regulations that allowed Tesla to retain EV dominance in the US, in the face of Chinese competition.
Tesla was very popular in the Chinese market and globally, including in markets where Chinese EVs aren't banned, until literally this year, which I'd argue is due in part to the trade war.
The real whole shtick is run economy in closed cycle to keep currency weak. Or the good old 1930s trade bloc economy. They're not just good at optimizing costs, they charge appropriately in CNY and inappropriately in USD. Workers don't care about obscene undervaluation in USD so long that they have bacon on the table after few hours of work.
It's not that rare that Chinese products are sold below cumulative costs of Western equivalent products and services, let alone prices. Chinese(<-substitute this with appropriate East Asian nations past and future) economy just isn't coupled well with the rest of the world that USD converted cost calculations would work. This in economic theories is sometimes explained as exports of starvation and/or overproduction, but IMO that make less sense when they've been doing it at scale of multiple decades.
The craziest example of these is Chinese PCB prototyping services: as cheap as $2 per 5 pieces with $5 extra for complete assembly and $15 shipping. $5 each would be darn cheap in the rest of the world, even $50 each for the board and $150 per assembly work would not be so absurd. There's just no competing that.
> I wonder why nobody has tried to beat the Chinese companies at their own game. The whole schtick is: take a product that people like, vertically integrate and drive down costs. This is like the purest form of capitalism.
I think there are a lot of different reasons:
1. A lot of those Chinese competitors are involved in extremely intense cut-throat competition, which drives a lot of innovation that benefits a lot of stakeholders except investors (IIRC the term is "involution"). The the US, the investors a almost literal kings and their returns are paramount, and they'll even throw their own country under the bus if it means their returns are higher.
2. The US (in-general) has been letting its manufacturing capabilities wither for decades, while China has been building them up. Even if you wanted to beat the Chinese companies at their own game, the skills, suppliers, and scale to do that aren't available in the US anymore.
3. Working conditions in China are atrocious and pay is lower, which really helps if you're trying to undercut on cost.
The promise didn't pan out for us. You have to prepare and cordon off the floor, and the unit gets stuck half the time. Somehow it's exactly the right height to get wedged under furniture.
My wife got one to try and automate away the vacuuming. We went through the same thing, and it still needs to be babysit anyways. For the sheer amount of time and frustration of basically all robot vacuums, it has been easier to just get a nice Kenmore upright bagged vacuum and do it yourself anyways (and the results are basically always way nicer).
We've had the Chinese dreame with valetudo (so fully disconnected from the cloud), it works honestly very well. Doesn't get stuck, avoids my son's toys rather effectively and just works. I run it when we leave home.
There's honestly no reason why roomba didn't make something equivalent to the dreame, they could have competed with the Chinese manufacturers on feature and by allowing users to easily disconnect from the cloud. They didn't because the company was completely mismanaged and their products barely evolved.
I'm pretty optimistic about biped robots for this. Either you buy or lease one and a cleaning service teleoperates it, or subscribe to a cleaning service that drops it off and teleoperates for a few hours a week. Suppose it could even walk itself to the next customer if close enough.
If a wheeled low-profile vacuum that stays in a room is too hard to deliver, surely we can fix it by making it walk around and grasp a vacuum cleaner and walk between houses.
I remember the "upgrade for pets" option, which... didn't work. After buying the maxed out version I realized that the product simply had a long, long way to go - but iRobot did nothing with it other than launch new segments like "upgrade for vision based mapping" etc.
They could improve the design and get people to replace their machines with the improved ones, repeat and repeat.
Or they could sell the broken design and people would just buy more as they broke. They don't care if Costco was eating the cost with their in-house warranty.
The fundamental problem though is the same with all "household gadget" products. They look cool, and appear to solve a problem, but that is actually all a perception based on novelty. They actually don't work very well, they are not built very well, and they don't last very long. There's no point in improving them because the concept is fundamentally something people don't need in the first place.
Just buy a good canister vacuum and you're set for a decade or more. It will cost more than the latest gadget from Shark or Dyson or iRobot but it won't frustrate you and it will just reliably do what it is supposed to do without uploading anything to an IP address.
> Just buy a good canister vacuum and you're set for a decade or more. It will cost more than the latest gadget from Shark or Dyson or iRobot but it won't frustrate you and it will just reliably do what it is supposed to do without uploading anything to an IP address.
Cords suck. So I bought a cordless vacuum, and was able to vacuum more. But I also needed a mop because vacuums don't do well enough on my laminate, stuff still gets stuck on. So I bought a cordless mop, so I could map more. This worked great for awhile but...
But it turns out if I did my vacuuming and mopping every night, I could keep my floor in better condition. I don't have time for that, but a robot from Eufy does and doesn't cost much compared to how much I would benefit from it.
Luddism on HN is a bit weird, but I get it, some people don't see the point of automating these tasks because their lives aren't complicated enough yet (e.g. they don't have kids, or have lots of free time and energy to spend on house work).
I'm also a Miele canister vacuum owner, and everywhere in my house where I vacuum is within range of a wall outlet. When I'm done, the cord retracts into the vacuum so I don't need to wind it or stow it myself. I guess, for me, that takes care of the issue to a great enough extent that I just never saw an advantage that justified the expense?
If you are ok with it, I think that's fine. Cordless to me is a huge productivity boost since I can just pick it up and vacuum whenever. I think most people see it as a huge win, but I haven't conducted a formal poll or anything.
Having a robot do everything is just another step in the convenience direction. It is great if you have expensive floors that you want to maintain on a daily or bi-daily basis.
Cordless is nice until the batteries won't charge anymore. Or the charger stops working. Or you forgot to charge it and now want to use it. Or the charging connector gets worn and unreliable. Then you have an expensive battery replacement or other repair or (more likely) you just replace the whole device because it was made to be unrepairable, and now you have several pounds of plastic and e-waste to dispose of.
Dealing with plugging a cord into an outlet is no more burdensome than picking up the socks or shoes before the Roomba wakes up and tries to ingest them.
If the batteries don't work anymore, I buy a new vaccum. My Dyson was last updated in 2020, it is 2025 now, so I think it is working out? The charging dock works great for not forgetting to replace it.
I guess this is how people felt when they moved from wired phones to wireless phones?
> Dealing with plugging a cord into an outlet is no more burdensome than picking up the socks or shoes before the Roomba wakes up and tries to ingest them.
And dragging the cord around, and having to plug out and re-plug the cord in again because you want to do a different part of the room.
> They could improve the design and get people to replace their machines with the improved ones, repeat and repeat.
> Or they could sell the broken design and people would just buy more as they broke. They don't care if Costco was eating the cost with their in-house warranty.
This strategy has limits, and I think iRobot hit those, and they didn't didn't lower themselves to switch to the second strategy of selling cheap unreliable garbage (at least not before 2019, which was the last time I bought a Roomba).
> The fundamental problem though is the same with all "household gadget" products. They look cool, and appear to solve a problem, but that is actually all a perception based on novelty. They actually don't work very well, they are not built very well, and they don't last very long. There's no point in improving them because the concept is fundamentally something people don't need in the first place.
I'd dispute this in this case: Roombas may not have solved the vacuuming problem for everyone, but they solved it for me (at least), and they were built pretty well (reliable, modular & reparable design, etc.).
> Just buy a good canister vacuum and you're set for a decade or more. It will cost more than the latest gadget from Shark or Dyson or iRobot but it won't frustrate you and it will just reliably do what it is supposed to do without uploading anything to an IP address.
1. I've got both, and the Roomba works a lot better than not vacuuming with the canister vacuum at all. It doesn't frustrate me, and it took far less time to Roomba-proof my home than vacuuming it every week for a year.
2. I agree with the IP address thing, but I think at only got added when they attempted to "get people to replace their machines with the improved ones." I have a couple of the older models that have no network connection (and had no plans to buy more due to the unnecessary network requirement).
They were saying that whoever was running things at Roomba must have been duped by the 4 hour work week bs because nothing was getting done. Specifically whoever took over operations, planning, and product improvements from Brooks.
Why not? Unless you are a Chinese citizen, it arguably makes more sense to grant access to the Chinese government rather than the US government. The PRC generally shows little interest in non-citizens while the US government frequently goes after people beyond its borders (e.g. Meng Wanzhou, Changpeng Zhao, Sam Bankman-Fried, Julian Assange, Kim Dotcom, etc.).
> Why not? Unless you are a Chinese citizen, it arguably makes more sense to grant access to the Chinese government rather than the US government.
You're making zero sense:
1. I predict there will be no change in the US government's access as a result of this.
2. I don't think Americans are so indifferent to their own country that they'd prefer a situation where an adversary country gets handed an intelligence asset. I mean, hypothetically, would an American prefer US trade policy be set that in a way that disadvantages American workers, because some politician got blackmailed because of something his Roomba recorded?
> The PRC generally shows little interest in non-citizens while the US government frequently goes after people beyond its borders
3. The Chinese government has been going after people in the US. They've long been engaged in industrial espionage, but there's also their "overseas police stations" (https://www.bbc.com/news/world-us-canada-65305415). It's worth noting that US citizens can have a Chinese origin, and I doubt the Chinese government would suddenly become uninterested in a dissident once he got naturalized.
I'm not sure how useful of an intelligence asset a map of my house is, or pictures of me in my boxers on a saturday morning. I'm also not sure why you think they weren't already just buying this information from iRobot.
> I'm not sure how useful of an intelligence asset a map of my house is, or pictures of me in my boxers on a saturday morning.
Seriously, who cares about you or your house? Why do you think your personal example is the one to reason from?
iRobot sold 50 million robots, lets conservatively say 10% of those are internet connected and still in service. That's 5 million households. There's probably quite a few people in that 5 million who have something going on that Chinese intelligence is interested in, even things that may affect you personally, if indrectly (if that's what you care about).
> Again, I don't know why you think China wasn't already buying that information from iRobot before it went bankrupt.
Come on, do really you think iRobot they sold data like that to third parties? Like user-identified floor plans? Camera images from inside people's homes?
> Aside from attempting to subvert democracies with botfarmed divisive politics, sure.
When Twitter had its recent VPN reveal, what actually took me by surprise was how many divisive accounts weren't from China or Russia, but from regions of the world like Turkey, India, Africa, and South America. Sure, they could be spouting divisive politics to push an agenda of someone who is paying them, but the simpler answer might be that they spout divisive politics because it earns them money in terms of advertising dollars.
And that's the real problem, IMHO. The subversion of democracy isn't happening because of China, Russia, or any number of adversarial countries, but because our social media companies don't care enough about our country or the people living inside it to meaningfully crack down on ragebait engagement farming.
The general thing about state actors is that they have every incentive to have a dossier of compromising information on every foreign national regardless of current relevance, for potential use in the future. You could, for instance, someday be in a position where you have privileged access to data that becomes relevant to them, and thus your history becomes useful.
True, but the US has a long track record of pursuing both foreigners and citizens—through prosecutions, extraditions, sanctions, or asset seizures—often years later and regardless of nationality. In practical terms, the risk of being targeted by the US for breaking a law is far higher than being blackmailed by a foreign state like China. The consequences are asymmetric as well: blackmail usually amounts to little more than embarrassment, whereas being pursued by the US government can carry lengthy prison sentences or worse.
This is not to mention that the US also engages in data collection for coercion purposes.
State actors also have finite budgets and do cost/benefit analyses. They don't really gain much for creating and maintaining deep dossiers on hundreds of millions of random foreigners.
and then list out examples that the US targeted which includes a US citizen, also while ignoring that China goes beyond its borders to target their citizens?
Both countries routinely act this way because they have the power to do so
Since this can be a significant security issue for the state, why doesn't the government sponsor a security audit of the software. Does it upload the data or everything is done on the device? (Also, will have to keep up with the updates)
Because regulation is bad, according to the current executive?
Politics aside, the FDA applies a very generous amount of regulation (mostly justifiable), not sure we want to pay multiples for our consumer electronics, as it (mostly) shows acceptable behavior and rearely kills anybody.
It is bad. Regulations have been historically hijacked to benefit corporate interests. See Intuit and tax policy for example.
Voters on the right naively thought he'd work to fix it. (Wrong!) But it is very much bad for a very large number of issues. Maybe next executive will fix it? (Wrong!)
The NSA has a bad historical reputation for this sort of thing - intentionally weakening crypto standards to make things easier for themselves to break, while keeping them "strong enough" that other agencies outside of NSA/GCHQ/GRU can't. The Crypto AG scandal [0] was pretty bad, with Clipper/Skipjack & Dual_EC_DRBG [1] being more recent ones. The NSA could do what you are asking to do, but they probably won't let us know what the really bad holes are because they want to keep using them.
Lol was thinking the same. Btw if you're into sensors, keen to chat. I'm the founder of crewline.ai, just raised our seed and are bringing on a couple key founding engineering hires.
The inside of a lawmaker's house? A general's? A CEO's? Why would anyone ever want insider information, including possible blackmail evidence, from them?
People like that have a cleaning service, they don't do their own housework. And ironically, the cleaning service is far better equipped to plant surveillance devices or exfiltrate information than a robot vacuum gadget.
Also I would assume it's a lot more dangerous and expensive to send someone in when you can just put an innocuous robot into a room which has cameras and microphones that can watch + listen 24/7 and auto-recharges when the battery is low (unlike surveillance devices).
Let's just say a family member has a very important and somewhat secretive job. The most classified meetings don't happen in our home, of course, but for a state adversary even small clues can be what's needed.
(So why get a roving camera in the first place? We judged that one from a historically and currently aligned state would be safe enough, even though it's not ideal.)
Blackmail for one: "I'll wage a campaign to flood the Internet with pics of your half dressed wife unless you....". That will work on some people, discovering extramarital affairs, or proving you know intimate details about the inside of someone's house, along with threat of physical violence will work on others. You just need to be sufficiently creative. You can parlay the successes from one target to the next: "I can get both of you divorced unless you install the rootkit on your work's network"
I will reference a quote I originally heard on HN years ago, though: the audio surveillance is magnitudes more valuable than the video.
Maybe I'm just not paranoid enough, but when people are having sexy time, they are not really likely to be running the roomba. Having that thing running is the opposite of sexy. When the thing is parked on the base station, it is facing the wall. So exactly how frequently do people think a roomba is running to be capturing all of this explicit footage?
To be clear, I'm not saying footage can't be captured, but some of these examples are just bat shit crazy well beyond paranoid
Yes, that's deliberately missing the point of what was said though. We know the vacuum is using the camera (when equipped) while the unit is running. Go back and re-read what I wrote to see why this is not the same thing.
> they are not really likely to be running the roomba.
You seem to assume that they have somehow physically disabled access to any kind of remote activation. That seems extremely unlikely given the overall selling points of the roomba.
The roomba doesn't have to "run" in order to be using its microphone, which as noted is likely the more valuable data acquisition source here.
Maybe read the comment I was replying to before replying with something unrealted? "I'll wage a campaign to flood the Internet with pics of your half dressed wife unless you....".
Turning on a microphone is not going to capture the required pics for this blackmail campaign.
Its a numbers game, and what I described is just a sketch. Combine it with geo-location: select all roombas within 50mi of washington, DC and run all these various playbooks on it, eventually you'll find something.
Why would the Chinese government want to regularly launch cyber attacks against US infrastructure, except it's been happening for years? US security companies and governments have been defending against it for years and have even cataloged the state-sponsored attack groups.
I don't like this kind of surveillance any more than anyone else on HN, but we get this here all the time. People make these posts that X leads to Y and jump way over the details. Sometimes X does lead to Y. Other times, it's the Underpants Gnomes: There's a big "???" step between X and Y that people don't like to take the time to articulate. This is how conspiracy theories take hold--you ignore the ??? and just assume "Of course X leads to Y! We all know it!"
HN should be above that. When we make a claim that X leads to Y we should be ready to show how X leads to X1, which leads to X2, which leads to X3, which leads to Y.
Almost all articles in the press about data collection and privacy are very poor and only focus on what data gets collected, not how it's used, nor how the circle completes and it comes back to harm the source of that data. To its credit, your second link at least lists a single vague example of how it's used, "data can be misused in ways such as fraudulent insurance claims or fake medical histories" but nothing about how that results in harm to the end user. We should expect better from reporters.
We should expect better from HN though, too. Let's not make conspiratorial claims here. I'm going to call them out, even though I am an opponent of this kind of data collection, too.
Yes? You also have several cameras and microphones in your pocket, and probably wearing more microphones on your head at various times of the day. And so does everyone around you.
GP is correct. "Roving camera/microphone" -> ??? -> "harm". What is "???" and what is "harm" specifically, and how the former leads to the latter, in specific steps?
Why are all us HN old timers in here arguing about this
I thought this was settled
People are walking around with self spy devices and putting them everywhere and giving all their private data to corpos.
That’s not new, we know it happens, we know companies use “anonymized” data for advertising. Its in public records for large companies balance sheets and there are thousands of data brokers who live exclusively on this data.
There are multiple compelling and popular documentaries about this.
Ironically we bought a Roborock (Chinese brand with close links to Xiaomi) and didn't connect it to the internet (checked this would work before we bought it).
If you don't want the scheduling and other app features, and are happy switching it on when you need it, it works fine.
Motivation for an offline one was more than just cameras, also that it wouldn't be bricked by an update one day, but still...
What is not for sale today, can be for sale tomorrow. Even Apple and Alphabet, should their leaders see greater market value in selling data rather than not selling it.
Well, I hope you learned your lesson and now won't blindly trust any corporation again, but rather demand open code and full control over the device you bought.
Especially if it is a moving camera in your home...
At least in the US, the Chinese government realistically should have been the least of your worries. What's China gonna do, if they caught you reading the Quran, or snorting crack? They could livestream your marriage proposal in WeChat to a billion people and you wouldn't ever notice. Meanwhile Snowden revealed, covertly watching random people through webcams is a leisure activity at the NSA, a national institution incidentally sharing jurisdiction with you. And evidently, your wife's death in a car accident may become a trending video at Tesla headquarters, while they deny your claims for a lack of such evidence.
> demand open code and full control over the device you bought
What do we do with companies/products like Tesla, short of shutting them down? Fully open code and absolute full control seems like it's going too far. Idealistically I like it, but practically I can't see it working.
I was making a point about trust, not what to do about it. If you are American, you should worry about Americans spying on you, as that actually could have consequences for your life and there is well know evidence and a legal foundation to justify such worries.
As non-car owner, I also dislike the Tesla cameras around me. Maybe one solution would be to not have fucking cameras everywhere, if the owner's exclusive access can't be guaranteed, abuse can't be prevented and legal consequences are not enforced. Maybe there should be standards and certification.
They were so proud of how they were one of the early ones to move their entire manufacturing process to China. Sorry, they always (could have) had a camera in your home
Lol was thinking the same. Btw if you're into sensors, keen to chat. I'm the founder of crewline.ai, just raised our seed and are bringing on a couple key founding engineering hires.
Back when Amazon was going to buy Roomba so they could use the cameras to sell us crap and/or sell the feed to law enforcement, I unplugged ours.
They were unimaginably unreliable compared to our older Roombas, and I was kind of shocked how little we missed them.
Anyway, I looked into getting a secure (or at least not malicious) alternative. At the time, the best bet was to get a Chinese model, then MITM its connection to the cloud + run your own server locally.
At that point, I realized it was less effort to just manually vacuum the house and moved on. I'm certainly not the only one, given the size of the modder community for the Chinese competitors.
Now, I wonder how far the modders are from buying a handful of commodity components + just 3d printing the rest of the robot, since that's less effort than dealing with enshittification.
At this point I trust the Chinese government way more than almost every US tech giant.
I don't own a "smart" speaker. I've never liked the idea of having an always-on cloud-connected microphone in my house. Like, it's just asking for trouble. I don't necessarily assign malicious intent here. It's just a recipe for disaster.
But if you made me choose between an Amazon or Meta "smart" speaker and a Huawei speaker, I'm choosing Huawei.
As for robot vacuums, I don't see a reason they need to have a microphone. I wouldn't want one that did. I think I'd also prefer they had a LIDAR rather than a camera too but I can see that cameras can do things that LIDAR can't.
Anyway, I find these deep distrust of the Chinese government to be very... selective, given what our own governments are doing and I'm sorry but our tech giants are out of control.
Lol agreed. Btw if you're into sensor, keen to chat. I'm the founder of crewline.ai, just raised our seed and are bringing on a couple key founding engineering hires.
This reminds me of the line from "Jackie Brown": "You can't trust Melanie. But you can trust Melanie to be Melanie".
I owned an early Roomba an it would just bump into things and "bounce" off. There was some sort of rudimentary fencing devices you could use to keep it in an area. I guess they decided cameras and things work better but I feel like the original worked well enough. You still had to vacuum but especially with pets it kept the disorder under control.
I agree. I think people underestimate the size of the Korean Brood War scene, even relatively early on. In my country, I had seen some huge LAN parties with associated competitions, but then I got introduced to Korean Brood War competitions; they were filling stadiums with audiences and had pyrotechnics and professional TV productions and everything. It was insane.
I don't think this is true. Granted, last time I tried to get good at an RTS was toward the end of the Brood War era but the established wisdom at that time was very clear that hour-for-hour, time spent practicing resource management was much more effective than time spent practicing clicking quickly.
Yes, really good players click fast, but they also have impeccable resource management. The group I played with did run the obvious experiment: the best one of us was forced to play against the rest (one at a time) with an artificial click frequency limit. He felt like his abilities were greatly reduced, but he still beat everyone else quite easily.
Yeah, I played a lot of StarCraft 2. By myself, 2v2 with a really talented friend and 3v3 with two other friends that were total beginners that I could beat 1v2.
At the bottom to upper mid level all you need to win is to figure out the macro game of building construction while also getting enough workers and units. With enough of that no micro is needed, just attack-moving into the enemy is more than enough.
Then at the upper mid level you're going to run into people who often don't build as effectively but they'll micro every unit or they'll be constantly doing raids when you don't expect it, scouting better than you and/or just understanding which units are better vs which so as to counter you.
From that point on it becomes much more of an effort to play the game because then you need to become better in all of those fields, while also becoming faster. But to be honest that point is probably 2/3rd's up the tree of all the people playing.
When people complain about APM in an RTS like StarCraft, they’re really not complaining about the spam clicking done by players at the pro level. They’re talking about multitasking which is an essential skill at all levels of the game.
Not even at the lowest rankings are you permitted to ignore what your opponent is doing and focus on building workers and base facilities. StarCraft is infamous for the ability of anyone to sacrifice their economy to perform an early rush attack (most infamously with a ton of early zerglings).
To combat early rush attacks you need to be able to multitask: send out early scouts to see what your opponent is doing, if they have any hidden building on the map, how many workers they have, etc. You need to be able to do this while building your own workers, base facilities, and units for defence. This is the multitasking that so many struggle with and it’s required to be able to play at the most basic level!
Optimally queueing SCVs and marines and supply depots requires an APM of 11 or so in the early stages of a Brood War game on the fastest setting. Add a couple more APM for scouting, and we'ree still not talking crazy levels of multitasking.
Dealing with your opponent is a fact of every strategy game!
And yet if you watch low level players they’ll be fine with that until a bunch of zerglings show up at their base and then they panic trying to micro marines and repair bunkers while their minerals shoot up to 1000 and then they have no units and lose.
Keeping a scouting SCV alive in your opponent’s base while building more SCVs at home, building more barracks, building supply depots, killing the enemy scouting worker, and actually reading and correctly interpreting what your opponent is doing is non-trivial.
Their annual geometric mean return is 45 %! That's some serious overbetting. In a market that didn't accidentally align with their biases, they would have lost money very quickly.
I'm extremely skeptical of any attempt to prevent leakage of future results to LLMs evaluated on backtesting. Both because this has beet shown in the literature to be difficult, and because I personally found it very difficult when working with LLMs for forecasting.
The one place where radians are more convenient is when you are at the centre of the circle. Then something which is as wide (or tall) as it is far away subtends one radian in your view. (And correspondingly, if it subtends half a radian it is half as wide as it is far away, etc.)
This happens to be the most common situation in which I measure angles.
More convenient than degrees. This is unrelated to pi vs tau (using tau or pi doesn't change the meaning of radians, the properties you mention are not affected). What OP is getting at is that the same number of radians, e.g. 1.57 (quarter turn) is more naturally expressed as tau/4 than pi/2.
reply