To be fair, LLMs usually use em-dashes correctly, whereas I think this document misuses them more often than not. For example:
> This can be extraordinarily powerful for summarizing documents — or of answering more specific questions of a large document like a datasheet or specification.
That dash shouldn't be there. That's not a parenthetical clause, that's an element in a list separated by "or." You can just remove the dash and the sentence becomes more correct.
I don't know whether that use of the em-dash is grammatically correct, but I've seen enough native English writers use it like that. One example is Philip K Dick.
Perhaps you have—or perhaps you've seen this construction instead, where (despite also using "or") the phrase on the other side of the dash is properly parenthetical and has its own subject.
I think you're thinking of british-style "en-dashes" – which is often used for something that could have been separated by brackets but do have a space either side – rather than "em" dashes. They can also be used in a similar place as a colon – that is to separate two parts of a single sentence.
British users regularly use that sort of construct with "-" hyphens, simply because they're pretty much the same and a whole lot easier to type on a keyboard.
You can stop LLMs from using em-dashes by just telling it to "never use em-dashes". This same type of prompt engineering works to mitigate almost every sign of AI-generated writing, which is one reason why AI writing heuristics/detectors can never be fully reliable.
I guess, but if even in you set aside any obvious tells, pretty much all expository writing out of an LLM still reads like pablum without any real conviction or tons of hedges against observed opinions.
"lack of conviction" would be a useful LLM metric.
I ran a test for a potential blog post where I take every indicator of AI writing and tell the LLM "don't do any of these" and resulted in high school AP English quality writing. Which could be considered a lack of conviction level of writing.
There was a comment recently by HN's most enthusiastic LLM cheerleader, Simon Willison, that I stopped reading almost immediately (before seeing who posted it), because it exuded the slop stench of an LLM: https://news.ycombinator.com/item?id=46011877
> if you spend too much time interacting with LLMs, you eventually resemble one
Pretty much. I think people who care about reducing their children's exposure to screen time should probably take care to do the same for themselves wrt LLMs.
It reads exactly like all his writing over many years afaict. Which is to say - it reads well. Just because someone is clear, thoughtful, and thorough, does not make them an AI. AI writing is actually quite different to this.
I don't know what to tell you: that really does not read like it was written by a LLM. You were perhaps set off by the very first sentence, which sounds like it was responding to a prompt?
Consider not sending your doorbell footage to accompany that has no respect for user privacy, and is now actively partnering with a police surveillance company.
And it's cheap! If you want JUST cameras it's pennies a day, I'm running a 2010 gaming PC with a dozen services (which would cost >200 bucks from the usual suspects) and it's still only 20 bucks a month.
Blue Iris is a fairly amazing camera hosting solution. I am continually impressed with the quality of the code and feature set. (disclaimer: $80/yr but we actually pay a bit less).
This f shameless pretention of doing something noble - barely helpful above normal practices btw. - while manipulating clueless users into turning on mass-surveillance is revolting and disgusting. And ordinary employees figured this out, phrased, created content, implemmented, pubished, and are maintaining this dirty practice. Many times with (very misplaced) pride. Shame on all of them actively participating in this coward scheme!
I have a Ring. I got the email notifying me and was about to go disable it, but it doesn’t share anything. It says it will notify you that your camera detected the dog and then YOU choose to share the video or not. So I left it enabled, as it becomes a later choice. Effectively I’m not opted into sharing was my take.
But you’re opted into the automatic detection, regardless. Ring is still processing your video still to see if there’s a match to any Search Party unless you turn the feature off.
At what point will the police request a warrant to run their own Search Party without consent?
So what? It already processes my video to detect people walking toward my house. Thats why I bought a smart camera not a dumb camera
A dog isn’t even on par to identifying humans. But let’s suppose if there’s a shooter on the run and my camera detects it, would I share that. Yes of course. Why wouldn’t I?
These kinds of systems tends to also be used by abusive policemen monitoring their exes, and police tracking members of undesirable (but legal) activist groups, etc.
See, I have no problem with searches that involve warrants and probably cause. They could already violate the shit out of your privacy with a warrant. That's kind of the point of a warrant.
What is an equivalently capable alternative that puts the user more in control?
The reason Ring is popular isn't just marketing or network effect, it's that it works. Before Ring and clones, security camera / DVR combos were really hard to make effective, I tried. Maybe you'd have a totally reliable system with good video, but it'd fail to notify you when you need it to, or notify way too often. Battery power was infeasible because cameras couldn't sleep. Phone notifications were DIY. A long compounding list of things could go wrong and make you miss an important event or fail to record it entirely. I'm hoping those have caught up by now, but haven't found any.
Someone tried to enter by the window next to my front porch door, it prompted me to install a video doorbell because the dude that frightened my wife only received a ticket because there was no proof he tried (and fail) to force open a window and did not enter the house. If I had a recording he would have spent the night at the station and it would have been a criminal offense instead of a civil one.
Even if it only provides deterrence, and a slight chance of after the fact punishment, I don't feel idiotic for buying a "doorstep Webcam", the door is visible from the street so there is no expectation of privacy and I really don't care that someone else could access those recordings.
If I had indoors cameras they would be in a private network. But for a front porch camera the easiest to install IOT junk is perfectly serviceable.
This. Local youths started playing ding-dong-ditch last summer. It was harmless when it was 8-9pm on a Friday or Saturday. They escalated to banging on windows, then to doing it at 1am, and finally to damaging our garage door to the point it won't open.
My wife is extremely upset about all of this, and I'm not going to be bullied out of the opinion that 24/7 cameras are actually a good thing.
No, I was pissed that the only punishment, for a know offender, was a small fine.
The police knew the guy (young adults with bright orange hoodie are quite uncommon here) and they told me that he already did this a few time before moving in my neighborhood and that they never had enough evidences to do something else than fine him.
Also I think the police are bored in my city because there were 4 patrols cars in the street when I got back home.
Thanks for the heads up, just went in and disabled it on my 2 cameras. Next step will be to throw these privacy invading pieces of junk in the trash. Just need to find a comparable product.
Are there any wireless (running power to these locations is not an option) doorbell cams that record to local storage instead of the cloud? I refuse to pay a subscription for these things.
Ideally they would record to my server instead of onboard SD card so that the footage can't just walk away if someone grabs the camera.
Reolink has some good options in this area. Both wired and wireless security and doorbell cameras with microSD storage by default, but FTP and SFTP can also be used. I've been happy with my (wired) doorbell camera with just a microSD card.
I agree it's the same as "opt-out" but I like the phrase "opt-in by default" because it implies there's an affirmative "I want to participate in this" option somewhere, and that it is set to "yes" by default.
IMO it properly reflects that what looks like an active affirmative choice by the user is actually not.
I bought a Ring doorbell several years ago, because I wanted an app-connected doorbell camera. That's it.
Once Amazon started rolling out other stuff and it was clear they were setting up a private panopticon, I trashed it and went with Blink, adding five wireless cameras as well. Blink isn't much better, but they're not as broadly compatible and don't have as many bells and whistles yet - but they're getting there.
The next step is to roll my own stuff. I expect I'll have that done by the end of the year.
I'll also note that none of my Blink cameras are pointing outward from my home. The closest to that is that I have one mounted on the front corner eave, but I made sure to point it so that it's looking at where I park my vehicles, and turned off motion detection for the small area of the street that it can see.
What I don't understand is that most of the ones I've seen others install are mounted under the eaves of the house, pointing outward. What's the purpose of that? They're not going to capture anyone actually trying to mess with your property. Most of mine are mounted in trees (front yard) or on posts that I installed for that purpose (back yard).
When a man murdered a woman in front of my house last year, our Ring camera's photos of his car led to his arrest within 24 hours, so not entirely useless?
I think a good thought experiment to consider, in terms of defining what your own views are, is to consider that if every single person had a mandatory 24/7 uplinked camera on them with redundancies, then the number of unsolved crimes would rapidly approach zero. It would be essentially impossible to get away with crime, so the only crimes that would happen would be those of passion, ignorance, or the political elite who would certainly excuse themselves from such social obligations, as usual.
But I definitely would not want to live in that world. And I think that's true for most people. It's kind of interesting too because there's some really nasty arguments one can make about this like, 'What, you'd rather see children kidnapped and even killed than consenting to surveillance that won't even be looked at unless you're under suspicion?'
But it's quite disingenuous, because with any freedom there is always a cost, and that cost is often extreme. 40,000+ people die per year because of our freedom to drive, yet few would ever use that as an argument to prohibit driving.
I think it's the opposite. I think people would prefer the peace of mind of living in a high trust society. People like predictability and being able to trust people. I also think people would enjoy that laws that people pass are actually applied and we can efficiently apply the will of the people to the country.
>with any freedom there is always a cost
Laws ultimately would be what restrict your freedom, not the enforcement of them. I don't think freedom should rely on poor enforcement of laws.
One of the biggest 'culture shocks' I had when first moving to Asia, that eventually I'd see in numerous places, is visiting a food court during business hours. There's people shoulder to shoulder, so everybody goes to claim a table before getting their food. They do this by leaving various things, including their purses, on tables while they went to go queue up to get some food. There wasn't a camera anywhere. That is a high trust society, and it's amazing. What you're describing here is the opposite of a high trust society - when you have a camera on every person for fear they might do something bad, it's a 0, if not negative, trust society.
Your perceptions of other peoples' views are also off. Even with the current scope of government surveillance, 66% of Americans say that the potential risks outweigh the potential benefits. [1] And laws would not be what limit freedoms. Government and authority is not some abstract holistic entity. It's made up of people, like you and I. Would you feel comfortable with me being able to surveil every moment of your life? The difference between me and the person who would end up doing so is not this great gulf you might imagine.
For instance Snowden revealed that NSA officers would regularly collect and trade sexually explicit media obtained from surveillance. [2] They'd also use their position to spy on their love interests to the point it gained it's own little sardonic moniker 'LOVEINT'. The people that would be looking through those cameras are just people. And the government leadership overseeing these groups would include those prone to go off to an island to screw minors, or more upstanding fellows like Eric Swalwell, cheating on his wife with a Chinese spy while serving on the House Permanent Select Committee on Intelligence that oversees the entire US intelligence apparatus, and would oversee this sort of surveillance.
>What you're describing here is the opposite of a high trust society
In such a society, people don't steal because thieves have been removed from society. You can trust others because they have proven through their life that they are trustworthy. By trust I was talking between the people in society and not about the government trusting that people would not break laws. Humans are not perfect, so it's a bad assumption to assume that citizens will not break laws.
>66% of Americans say that the potential risks outweigh the potential benefits
The article you linked was about the current benefits. This is different than what I am talking about where laws are able to be effectively be enforced.
>Would you feel comfortable with me being able to surveil every moment of your life?
What do I get in return from you? Nothing? Then I have no reason to do so.
>NSA officers would regularly collect and trade sexually explicit media obtained from surveillance.
This should be made an instantly firable offense, like it is in the tech industry for accessing personal data of users. There should be alerting when such data is accessed to ensure that systems are not being abused.
The article was specifically and explicitly speaking of potential benefits, not present. The problem with systems made up of people is that you can't just have turtles all the way down. Let's say that retaining sexually explicit media of surveilled individuals is indeed grounds for instant dismissal. Who enforces this? Okay, the person above the people doing surveillance, who's somehow going through the entirety of said surveillance. And what happens when he is the one saving such media? Is it then overviewed by another person above him? And so on.
You can't really have endless redundancies and, at scale, this becomes even more true where the vast amount of data and processing becomes ever less viable to filter. And when you had 24/7 footage of everybody at every moment, that takes scale and ups it to an inconceivably vast level. More generally I think removing thieves from society, with extreme prejudice, is a far more pleasant path forward for everybody than treating everybody like a potential thief.
> In such a society, people don't steal because thieves have been removed from society. You can trust others because they have proven through their life that they are trustworthy.
This is a very "Star Trek" view of the future and even Star Trek repeatedly demonstrates that the perfect society is an illusion.
I don't think it's necessarily an illusion, but rather its repeatedly demonstrated that it requires measure that some might not like. In "Asia" (quotes as it feels odd to characterize an entire continent of dozens of different countries, yet there is often a widely shared ethos on many topics) criminals tend to be largely ostracized from society. And it's not just the criminals, but also their family and relations. This is probably somewhat more akin to what Western society was in the past with things like pillories where the punishment wasn't just the humiliation, but everyone knowing that you were a criminal.
Even the justice systems tend to be different. For instance in the US you're expected to plead not guilty, even if you're guilty - and then work from there. In "Asia" 'falsely' pleading not guilty is often seen as a lack of remorse and generally results in far harsher penalties than pleading guilty. And when one pleads guilty, they're often even required to do things like reenact their crime and generally cooperate with the government in every single way. It's just a very different approach towards criminality.
Another problem is that “crime” is a very flexible term and when you have a hammer everything looks like a nail. Also the legal system specially in the US likes to equate convictions with solved cases which is not always the same.
I’m ok with that as long as I, the camera owner, am choosing to hand over the footage. At best I can see some sort of watermarking to ensure that it’s legitimate.
There are legitimate reasons to want a camera either at your front door or surveilling your property. These can range from an increased sense of security to having documentation to support insurance claims, or even for watching wildlife. We installed our Ring camera after an ongoing string of nighttime car break-ins hit us and we had no direct proof of what happened for insurance. It was meant to be both a deterrent to that type of event and also for documentation if it happened again. There's also a pack of coyotes that lives in the woods near our house and occasionally eats our chickens. While that usage was more out of curiosity (if you have chickens, you're going to lose one from time to time), we were able to develop a sense of when that threat was higher.
I live on a bucolic cul-de-sac in a house that I've lived in since the mid 1970s. Most of the neighbors are the same. I never in my life expected a random person to drive down the street, drag a lady out of his trunk, chase her around the cul-de-sac, and stab her to death in front of my house. I never expected to find the body in the woods 40' from my side door. This is when I also learned that nobody comes to clean up after a crime like that and that if I didn't want pools of blood in front of my house and a 50' streak of it crossing the circle or the splatters all over the mailboxes that I was going to have to go out there and clean it up myself. I was in PTSD therapy for a while after that. I'm glad the Ring camera caught some of the activity.
After an event like that, it's easy to lose a sense of security in your home. How are you supposed to sleep the night after that happens, when the perpetrator remains at large? You can't lock your doors hard enough or do anything at all to feel secure. That lack of sense of security does not go away in a day or a week or a month. It goes away when you can find "normal" again. It helped us to find normal by installing other cameras around the house.
I don't want Ring or Arlo or anybody to be automatically sharing my camera footage with anybody. Even with the murder event, it was my choice to go through the footage and share it with the authorities. I don't support authoritarian "law enforcement" activities, I don't want anybody tapping into my camera feed to find lost pets or for any other reason. They shouldn't be allowed to do it. Like many other services we all use, we're more of the product than the customer, as our data is harvested and used for other purposes.
Personal security is different than targeted advertising. Most people won't know they need or want a camera until after they have experienced something that makes them feel less secure in their home. I just hope they have the wits to read the Terms and understand what they're opting into before automatically accepting all of the opt-in-by-default data sharing.
Not one disagreement with what you're saying. I have cameras outside my house. I'd like them to be end-to-end encrypted and am perfectly fine with a voluntary self-report feature. But what Ring seems to be pushing for is opt-out mass surveillance, and once connected to AI this means we're going to the bad place.
Even in the most dystopian sci-fi future where a hostile and totalitarian government watches everything everybody does, they would still use the information to investigate boring everyday crimes.
The (non rethoric) question is, are people willing to pay the increasing price of non-crime related surveillance as we see diminished security margins.
When a man murdered a woman in front of my house last year, our Ring camera's photos of his car led to his arrest within 24 hours, so not entirely useless?
Your doorbell photo of a car was really the only evidence to convict someone of murder?
I'm glad I live somewhere that needs more proof that.
No, it enabled them to find him quickly. There was other evidence, but with no previously know connection to the victim and the perpetrator having no prior criminal record, I was told it was unlikely they would have found him otherwise.
Just knowing when stuff happens outside your house when you're not home. Like, someone in the household came back from work, a package arrived, or a cool animal showed up. Or you're home and appreciate a little more notice before a visitor rings the doorbell.
Personally not a compelling enough reason to buy the camera in the first place, but those non crime notifications end up being the most common once it's up.
However, it's a very abbreviated way of stating, "soon you will not even be given this choice, because we make entirely too much money selling your data/info and we kinda bribe law enforcement by granting them any and all video, with a simple request"
The best time to not buy into all this non-free surveilence-as-a-service crap was a decade ago.
The second best time is today.
Unfortunately the public love this stuff, and are quite happy to have CCTV pointing at your house. Privacy never existed 300 years ago, it doesn't today. Accept your feudal masters and make peace with it, because they won years ago.
Is this seriously your conclusion? Might be a good time to get rid of the fucking spy camera owned by a multitrillion dollar corporation partnering with the state surveillance apparatus, is my opinion.
Have people never read/watched a sci-fi book/film before?
I think encouraging people to enable E2EE is more realistic than asking everyone to throw out the Ring cams they’ve potentially spent hundreds on, yeah.
I can see benefits to a closed circuit camera. I've never felt the need personally.
If it's not running free software and treated in a secure fashion (camera can't talk to anything other than the server, enforced at the network level, etc) then it's not a risk per-se.
However people would rather pay up front and pay subscriptions to get outside companies to run it than run their own their own equipment. In the 90s when I was excited about tech I didn't even consider that aspect.
From the linked document (in German for some reason, so I skimmed it as best I could), it sounds like the device will generate a password and you need to enter it on your phone. So symmetric encryption, not trusting the server to distribute keys that would be susceptible to MITM, and also not leaving users the options to choose bad passwords
Sounds good from a security point of view, although it also says they disable functions like having more than one person able to view the camera (having a partner be able to answer the door seems pretty fundamental; they probably just can't be arsed to make such functionality work with safety turned on...)
Of course, just like with Signal or anything else that gets regular updates, they can push an update to your device specifically that sends the decryption key out. You'll always have to trust them to not do something like that, but that's a whole different level from subpoenaing data they have on a hard drive
Is Ring camera encryption actually independently audited and known to be implemented correctly and provide all the desirable security properties?
Because when I reverse engineered my Tuya-based camera-equipped pet feeder, I've discovered that there was an encryption on the video stream, but they only encrypted I-frames and left P-frames unencrypted. Amazon is not Tuya, but IoT is IoT.
My point is, there are myriad of ways IoT vendor can boast "encryption" and "security" on the marketing materials, while the actual implementation could be flawed beyond redemption.
Just keep your cameras on separate vlan and access through eg. wireguard. Any company can have the best intentions but gov can just come to them, tell them to implement whatever is needed - even if that means lying to their users - and have access to everything. Probably even with plausible deniability for all parties involved, but not sure anyone even still cares about that.
And that’s the whole point - it’s APIs we did not have. Now app developers are encouraged to have a public, user friendly, fully functional API made for individual use, instead of locking them behind enterprise contracts and crippling usage limits.
I do have one: Atlassian now allows connecting their MCP server (Jira et al) for personal use with a simple OAuth redirect, where before you needed to request API keys via your org, which is something no admin would approve unless you were working specifically on internal tooling/integrations.
Another way to phrase it is that MCP normalizes individual users having access to APIs via their clients, vs the usual act of connecting two backend apps where the BE owns a service key.
Normal users don't know what MCP is and will never use an MCP server (knowingly or unknowingly) in their life. They use ChatGPT through the web UI or the mobile app, that's it.
MCP is for technical users.
(Maybe read the link you sent, it has nothing to do with defining a new standard)
Normal users will increasingly use MCP servers without even knowing they do so - it will be their apps. And having e.g. your music player or your email client light up in the ChatGPT app as something that you can tell it to automate is not just for technical users.
Isn't that what we had about 20 years ago (web 2.0) until they locked it all up (the APIs and feeds) again? ref: this video posted 18 years ago: https://www.youtube.com/watch?v=6gmP4nk0EOE
(Rewatching it in 2025, the part about "teaching the Machine" has a different connotation now.)
Maybe it's that the protocol is more universal than before, and they're opening things up more due to the current trends (AI/LLM vs web 2.0 i.e. creating site mashups for users)? If it follows the same trend then after a while it will become enshittified as well.
Tried it. It was okay, felt very futuristic. But ultimately there weren’t really any benefits over just using a multi monitor setup that could justify the cost. I thought it’d be nice to have something I could also use for watching movies, but when I watch movies I almost always watch them with my wife, and much prefer that.
reply