So many companies send emails with multiple different address. For example, Citibank could send from citicards@info6.citi.com or CustomerService@thankyou.citi.com. Bank of America can be communications@servicing.bankofamerica.com or onlinebanking@ealerts.bankofamerica.com. If the "Citibank inbox" just takes the first address that Citi happens to use and auto rejects all future others, user will lose critical emails without even knowing. There is no way either user or the app could guarantee an entity will ever only use one address for legitimate reason.
Ah, but Clova handles that. It locks to the domain, not the sender or the subdomain. So both your examples would still be allowed (citi.com and bankofamerica.com). There are also permanent inboxes in the system, which allow anyone. The trick is to delete them after a while when they get polluted with spam.
But Flock gives police overconfidence to commit actions instead of having due diligence. It is like the best student in your classroom tells you 2 + 2 = 5, and you naturally feel compelled to believe him, even there is clear evidence against the claim.
This is also comparable to self-driving car: Waymo's system has achieve the level of "less error than human on average" long time ago, but they are still only doing slow expansion. Why? Because when AI is involved, human naturally expect them to be almost 100% correct, not just barely better than human, otherwise why bother use AI (and all the troubles to legalize for AI) in the first place? So in this case, I guess the police just blindly trust the AI being 99.99% correct.
Flock isn’t an “AI” though? It doesn’t put together a case and say “yes, this is the criminal”. It’s not a crime-solving AI.
It’s a camera system, and data can be queried by police. They have to search for something. So the cops searched for a vehicle matching the description in the general area, seemed to take the first one, and called it a day.
That’s not an AI-uncanny-valley. That’s the equivalent of (pre-flock/cameras) looking at vehicle registration list and assuming the one at the top of the list must be it.
The automation-trust is a legitimate concern but it’s not applicable here.
I don't know the details on how Flock is internally designed, but I suppose AI is at least involved in the image recognition and correlation, at least to this scale. I also don't know exactly why Flock misidentifies Issacs' car in this case, because I don't find it in the article or the interview, but I assume it is due to a mislabeling of the plate number or the car damage imagery, which is related to AI. Correct me if I'm wrong.
Flock didn't misidentify Isaacs' car in this case. It correctly identified that her car was vaguely near the accident, and the police incorrectly concluded that it was the one involved in the accident.
Correct, but what's more, there is a lack of accountability issue that's part of it. When people are not held accountable for the screw ups, then processes don't or are very slow to change. That's how you can just call it a day, regardless of the harm done to innocent citizens.
Flock is not an AI. That they have AI features on certain products does not mean "Flock" is an AI. That's the fallacy of composition.
Flock does have AI components (this one seems LLM based). Basically every ANPR system will also use some level of AI.
Also, on "reality would like a word". Why not just speak normally instead of using what seems like a high-school grade comeback?
The main Flock related issue here is that officers treat results that come out of the systems as gospel. The same thing has happened with face recognition systems. There is also the broader issue of access controls to Flock, 5A, and whether places should be building up a technology so powerful (surveillance infrastructure can easily be used for evil).
I said that because your position is absurd, and clearly false. Flock is an AI company, period. They use a multitude of different ai types in their “AI OS”.
And the issue is absolutely not just cops taking what they say as gospel; that’s a teeny tiny itty bitty fraction of the issue, which is in itself reductive. This is a massively multifaceted issue and to state otherwise means you don’t understand the scope.
I have made no comments on this thread before the one above.
So saying you said something because my non-existent position is "absurd" and "clearly false" indicates massive levels of confusion.
You then go on to imply I haven't read the article, using an acronymised swear (I'm not a pious person but clearly directing a swear to someone makes a message come off as aggressive), even though your understanding of what I said is based off non-existent messages. (This is quite ironic).
Flock is a company. Companies are a type of organisation: it makes no sense to describe a company as AI, as AI is a type of computer system. It may make sense to say Flock is an AI-based company, but calling them "AI" in itself makes no sense.
I can't find anything indicating that an LLM was used in this case, and this county seems to mainly have Falcon cameras (https://www.scribd.com/document/983922612/250730110407Flock-...). Reverse engineering seems to indicate these are powered by non-LLM AIs like Tensorflow vision NNs. These have been around for ages, and notably no-one is alleging these models hallucinated etc.
People also (perhaps wrongly) use "Flock" to refer to the camera system. CrendKing explicitly calls "Flock" a "camera system". A system of cameras is not AI, and the point of the comment seems to be that it's inappropriate to bring up LLM related AI fears about this system.
> It is like the best student in your classroom tells you 2 + 2 = 5, and you naturally feel compelled to believe him, even there is clear evidence against the claim.
Yes, but it is still your fault. “I am dumb and I was deceived” is no valid defense, least of all for a police officer.
I'm not saying the police has no fault. Using AI assist without secondary verification is never good idea, at least not in 2026. But you can't argue Flock the technology itself, and especially Flock the company, are totally innocent, just like you can't say the "guns are just tools like hammer, it only depends on who uses them", and the gun industry.
It is not even clear from the article that there were anything wrong with the output. She has a car of that make. And it doesnt appear to be in dispute that she drove nearby.
It simply is the case that “being nearby” in itself is not enough evidence that you were involved in a crash. Especially when the colour doesn’t match and the vehicle is not damaged.
The fundamental problem here is that the police are very bad about false positives. Flock gives them a ton of data--but that doesn't make it magically perfect. No surprise there was a similar car nearby--there are a lot of similar cars. Flock shows them the similar cars and they pick the closest without any investigation.
> But you can't argue Flock the technology itself, and especially Flock the company, are totally innocent
For the sake of argument, let’s suppose that Flock is 100% evil/guilty/blameworthy in this case. Even in that case, two things are still true:
A) Flock has no authority to arrest anyone
B) Flock has no authority over the police
The buck has to stop somewhere. It’s counterproductive to focus on what Flock did, compared to what the police did. And I say that as someone who is not a fan of Flock.
So next time start your first post with that, otherwise you lead us on this wild goose chase only to find that you mostly agree, it's a foundational part of discourse.
I don't feel we are asking too much of police and lawyers here: yes, you can use technology and tools (such as Flock) to help you do your job more efficiently, but that doesn't mean you can turn your brain off.
Perhaps more facts will come out, but if we take the article at face value, the cops are at this woman's house, looking at her car and can see with their own eyes that the car right in front of them has obviously not been in an accident. All that is required is for these cops to rub two brain cells together and they will reach the conclusion that Flock misread a license plate.
It's also worth emphasizing that this wasn't for a parking ticket. The extremely low level of care with which these public servants are conducting a manslaughter investigation is somewhat alarming.
> It is like the best student in your classroom tells you 2 + 2 = 5, and you naturally feel compelled to believe him, even there is clear evidence against the claim.
It’s more like normal people don’t know that 97.8% accurate isn’t 100% - despite how good of a score that would be on a test.
Law enforcement is predominantly made up of normal people.
I would say US law enforcement are made up of mostly abnormal people with lower intellegences, higher aggressivness, and a propensity to lie and deceive. They tend to push actual normal people out for not playing ball with their own corruption.
The number one demograph for domestic violence and abuse are cops, they are not just a randomly selected group of people that chanced into law enforcement as a career.
> The number one demograph for domestic violence and abuse are cops, they are not just a randomly selected group of people that chanced into law enforcement as a career.
Be sure you don’t have the causation backwards - or that it doesn’t flow strongly both ways!
It attracts certain people - and being around that crowd and power pushes some number of people over the line that were near it.
See also: stimulant use, abuse, in the construction industry.
Absolutely, law enforcement in the US is disproportionately not normal people. The system is designed to reject smart people who are not aggressive and have empathy. The system is designed to accept people who will uphold the design of the system.
> It is like the best student in your classroom tells you 2 + 2 = 5, and you naturally feel compelled to believe him
Closer analogy: The dumbest student in a class tells you 2 + 2 = 5 and you believe them.
Even more apt analogy: You're the dumbest student in the class, the second-dumbest student tells you 2 + 2 = 5, but you lack the critical thinking skills necessary to know whether that student is smart or not so you just believe them.
I just tested. 7-Zip is already doing exactly what the author's recovery tool does: if I delete the metadata at the end of the file, it still can open and extract files, while complaining about the harmless "Unexepcet end of data". Further, even if I delete the actual content of the last file, it still can show the directory structure, it just can't extract data of the last file.
But how to you quantify the effect of the exposure? How do you know a particular sale is converted through some impressions you put on Facebook six months ago, and not from a highway billboard one week earlier? If you can't accurate quantify the effect of each ad campaign with their conversion rate, you have no reason to put more funds into this particular Facebook campaign. So naturally it falls on Facebook to prove to you that their ad socket is actually generating conversion, by not showing the ads to people who is least likely to click through ads when shown.
I think it's all a huge make-believe. They have a big name and marketing falls for it. If push comes to shove, they can just simulate click through and views and such, to "convince" people that it all works. They have all the user agent details of a user from years of browser-rape.
1) If it really was a bomb and went off, the pilot wouldn't be there to explain to the press anyway.
2) How likely would a bomb's name really be "BOMB" vs anything else? If the latter is any higher, wouldn't it be reasonable to always turn around whenever the any other name shows up? In that case, all Bluetooth devices should be strictly banned in the cabin. But TSA is not doing that (not yet).
Yesterday I found an app that I need to keep my bluetooth headphone from entering sleep mode. It is Sound Keeper: https://veg.by/en/projects/soundkeeper/. And it uses exactly the same approach talked in this article. For example, normally its filename is SoundKeeper64.exe. But if you rename it to SoundKeeper64AllOpenOnly.exe, it switches behavior to operate on "All" devices with the "OpenOnly" mode.
I believe most of the time this phrase is said to an inexperienced artisan who has no idea how the current system works, what's the shortcoming of it, and how to improve upon it. Think of an undergraduate student who tries to solve the Goldbach conjecture. Usually what ended up is either he fails to reinvent the wheel, or reinvent the exact same wheel, which has no value. The phrase certainly does not apply to professionals.
Even then, you know what's a good way to learn about how the current system works etc, maybe even the best way? I've got many failed projects behind me, and 0 regrets.
Why can't Cargo have a system like PyPI where library author uploads compiled binary (even with their specific flags) for each rust version/platform combination, and if said binary is missing for certain combination, fallback to local compile? Imagine `cargo publish` handle the compile+upload task, and crates.io be changed to also host binaries.
> Why can't Cargo have a system like PyPI where library author uploads compiled binary
Unless you have perfect reproducible builds, this is a security nightmare. Source code can be reviewed (and there are even projects to share databases of already reviewed Rust crates; IIRC, both Mozilla and Google have public repositories with their lists), but it's much harder to review a binary, unless you can reproducibly recreate it from the corresponding source code.
I don’t think it’s that much of a security nightmare: the basic trust assumption that people make about the packaging ecosystem (that they trust their upstreams) remains the same whether they pull source or binaries.
I think the bigger issues are probably stability and size: no stable ABI combined with Rust’s current release cadence means that every package would essentially need to be rebuilt every six weeks. That’s a lot of churn and a lot of extra index space.
> remains the same whether they pull source or binaries.
I don't think that's exactly true, it's definitely _easier_ to sneak something into a binary without people noticing than it is to sneak it into rust source, but there hasn't been an underhanded rust competition for a while so I guess it's hard to be objective about that.
Pretty much nobody does those two things at the same time:
- pulling dependencies with cargo
- auditing the source code of the dependencies they're building
You are either censoring and vetting everything or you're using dependencies from crates.io (ideally after you've done your due diligence on the crate), but should crates.io be compromised and inject malware in the crates' payload, I'm ready to bet nobody would notice for a long time.
I fully agree with GP that binary vs source code wouldn't change anything in practice.
> Pretty much nobody does those two things at the same time:
- pulling dependencies with cargo - auditing the source code of the dependencies they're building
Your “pretty much” is probably weaseling you out of any criticism here, but I fully disagree:
My IDE (rustrover) has “follow symbol” support, like every other IDE out there, and I regularly drill into code I’m calling in external crates. Like, just as often as my own code. I can’t imagine any other way of working: it’s important to read code you’re calling to understand it, regardless of whether it’s code made by someone else in the company, or someone else in the world.
My IDE’s search function shows all code from all crates in my dependencies. With everything equal regardless of whether it’s in my repo or not. It just subtly shades the external dependencies a slightly different color. I regularly look at a trait I need from another crate, and find implementations across my workspace and dependencies, including other crates and impls within the defining crate. Yes, this info is available on docs.rs but it’s 1000x easier to stay within my IDE, and the code itself is available right there inline, which is way more valuable than docs alone.
I think it’s insane to not read code you depend on.
Does this mean I’m “vetting” all the code I depend on? Of course not. But I’m regularly reading large chunks of it. And I suspect a large chunk of people work the way I do; There are a lot of eyeballs on public crates due to them being distributed as source, and this absolutely has a tangible impact on supply chain attacks.
> Does this mean I’m “vetting” all the code I depend on? Of course not.
Inspecting public facing parts of the code is one thing, finding nasty stuff obfuscated in a macro definition or in a Default or Debug implementation of a private type that nobody is ever going to check outside of auditors is a totally different thing.
> My IDE (rustrover) has “follow symbol” support
I don't know exactly how it works for RustRover, since I know Jetbrain has reimplemented some stuff on their own, but if it evaluates proc macros (like rust-analyzer) does, then by the time you step into the code it's too late, proc macros aren't sandboxed in any ways and your computer could be compromised already.
The point of my argument is not to say I’m vetting anything, but to say that there are tons of eyeballs on crates today, because of the fact that they are distributed as source and not a binary. It’s not a silver bullet but every little bit helps, every additional eyeball makes hiding things harder.
The original claim is that “pretty much no one” reads any of their dependencies, in order to support a claim that they should be distributed as binaries, meaning “if there was no source available at all in your IDE, it wouldn’t make a difference”, which is just a flatly wrong claim IMO.
A disagreement may be arising here about the definition of “audit” vs “reading” source code, but I’d argue it doesn’t matter for my point, which is that additional eyeballs matter for finding issues in dependencies, and seeing the source of your crates instead of a binary blob is essential for this.
> The original claim is that “pretty much no one” reads any of their dependencies,
No the claim is that very few people read the dependencies[1] enough to catch a malicious piece of code. And I stand by it. “Many eyeballs” is a much weaker guarantee when people are just doing “go to definition” from their code (for instance you're never gonna land on a build.rs file this way, yet they are likely the most critical piece of code when it comes to supply chain security).
[1] (on their machines, that is if you do that on github it doesn't count since you have no way to tell it's the same code)
> No the claim is that very few people read the dependencies[1] enough to catch a malicious piece of code.
You’re shifting around between reading enough to catch any issue (which I could easily do if a vulnerability was right there staring at me when I follow symbol) to catching all issues (like your comment about build.rs.) Please stick with one and avoid moving goal posts around.
There exists a category of dependency issues that I could easily spot in my everyday reading of my dependencies’ source code. It’s not all of them. Your claim is that I would spot zero of them, which is overly broad.
You’re also trying to turn this into a black-or-white issue, as if to say that if it isn’t perfect (ie. I don’t regularly look at build.rs), it isn’t worth anything, which is antithetical to good security. The more eyeballs the better, and the more opportunities to spot something awry, the better.
I'm not moving the goal post, a supply chain attack is an adversarial situation it is not about spotting an issue occurring at random, it is about spotting an issue specially crafted to avoid detection. So in practice you are either able to spot every kind of issues, or none of the relevant ones because if there's one kind that reliably slips through, then you can be certain that the attacker will focus on this kind and ignore the trivial to spot ones.
If anything, having access to the source code gives you an illusion of security, which is probably the worse place to be in.
The worse ecosystem when it comes to supply chain attacks is arguably the npm one, yet there anyone can see the source and there are almost two orders of magnitude more eyeballs.
In such an environment I’m doomed anyway, even if I’m vetting code. I don’t understand why the goal has to be “the ability to spot attacks specifically designed to prevent you from detecting.” For what you’re describing, there seems to be no hope at all.
It’s like if someone says “don’t pipe curl into bash to install software”, ok that may or may not be good advice. But then someone else says “yeah, I download the script first and give it a cursory glance to see what it’s doing”, wouldn’t you agree they’re marginally better off than the people who just do it blindly?
If not, maybe we just aren’t coming from any mutual shared experience. It seems flatly obvious to me that being able to read the code I’m running puts me in a better spot. Maybe we just fundamentally disagree.
> It’s like if someone says “don’t pipe curl into bash to install software”, ok that may or may not be good advice. But then someone else says “yeah, I download the script first and give it a cursory glance to see what it’s doing”, wouldn’t you agree they’re marginally better off than the people who just do it blindly?
I don't agree with your comparison, in this case it's more like downloading, then running it without having read it and then every once in a while look at a snippet containing a feature that interest you.
The comparison to “download the script and read it before you run it” would be to download the crate's repo, read it and then vendor the code you've read to use as a dependency, which is what I'd consider proper vetting (in this case the attacker would need to be much more sophisticated to avoid detection, it's still possible but in this case at least you've actually gained something), but it's a lot more work.
If you have reproducible builds it's no different. Without those binaries are a nightmare in that you can't easily link a given binary back to a given source snapshot. Deciding to trust my upstream is all well and good but if it's literally impossible to audit them that's not a good situation to be in.
I think it’s already probably a mistake to think that a source distribution consistently references a unique upstream source repository state; I don't believe the crate distribution layout guarantees this.
(I agree that source is easier to review and establish trust in; the observation is that once you read the upstream source you’re in the same state regarding distributors, since build and source distributions both modify the source layout.)
It might as well. If there is no definition of an ABI, nobody is going to build the tooling and infrastructure to detect ABI compatibility between releases and leverage that for the off-chance that e.g. 2 out of 10 successive Rust releases are ABI compatible.
You can have binary dependencies with a stable ABI; they're called C-compatible shared libs, provided by your system package manager. And Cargo can host *-sys packages that define Rust bindings to these shared libs. Yes, you give up on memory safety across modules, but that's what things like the WASM Components proposals are for. It's a whole other issue that has very little to do with ensuring safety within a single build.
docs.rs is just barely viable because it only has to build crates once (for one set of features, one target platform etc.).
What you propose would 1) have to build each create for at least the 8 Tier 1 targets, if not also the 91 Tier 2 targets. That would be either 8 or 99 binaries already.
Then consider that it's difficult to anticipate which feature combinations a user will need. For example, the tokio crate has 14 features [1]. Any combination of 14 different features gives 2^14 = 16384 possible configurations that would all need to be built. Now to be fair, these feature choices are not completely independent, e.g. the "full" feature selects a bunch of other features. Taking these options out, I'm guessing that we will end up with (ballpark) 5000 reasonable configurations. Multiply that by the number of build targets, and we will need to build either 40000 (Tier 1 only) or 495000 binaries for just this one crate.
Now consider on top that the interface of dependency crates can change between versions, so the tokio crate would either have to pin exact dependency versions (which would be DLL hell and therefore version locking is not commonly used for Rust libraries) or otherwise we need to build the tokio crate separately for each dependency version change that is ABI-incompatible somewhere. But even without that, storing tens of thousands of compiled variants is very clearly untenable.
Rust has very clearly chosen the path of "pay only for what you use", which is why all these library features exist in the first place. But because they do, offering prebuilt artifacts is not viable at scale.
You could get a lot of benefit from a much smaller subset. For example, just the "syn" crate with all features enabled on tier 1 targets (so ~8 builds total) would probably save a decent chunk off almost everybody's build.
It runs counter to Cargos curreat model where the top-level workspace has complete control over compilation, including dependencies and compiler flags. I've been floating an idea of "opaque dependencies" that are like python depending on C libraries or a C++ library dependening on a dynamic library.
That would work for debug builds (and that's something that I would appreciate) but not for release, as most of the time you want to compile for the exact CPU you're targeting not just for say “x86 Linux” to make sure your code is optimized properly using SIMD instructions.
Maybe I'm dumb, but I can't see how the code in the "Errors and map" section can compile. "transform_list" returns a Result<>, yet "result" is just a Vec. I thought you always need to wrap it with Ok()? Is that a new nightly feature?