Hacker News Reader

> Look, I’m not trying to be pro-surveillance state or whatever

Kinda sounds like you are, though.

> but in what universe is a handful of cameras (less than 10 per building)

The article notes 6-7 cameras per floor of one particular building. I'm not sure of the size of each floor of that particular building, so 6-7 cameras per floor may cover very little space, or it may cover a lot. That also says nothing about the coverage in other buildings. Larger buildings probably got more cameras.

Either way, often these sorts of things are set up in phases, with higher concentrations of cameras added later, after the system proves itself as useful.

> My high school also had cameras. Isn’t that rather typical?

There's a big difference between your bog-standard security camera that just records footage to a hard drive somewhere, and these cameras, that send live feeds to an AI company, where the video is presumably automatically analyzed and indexed to allow administrators to effortlessly track people's movements. The AI company may (now or in the future) offer real-time alerts of "suspicious activity", which as we all should know by now probably will have a significant false-positive rate. (But cops will treat alerts as gospel without doing any critical thinking.)

There were a lot of things in the article I didn’t take at face value due to the obviously non-neutral point of view of the faculty.

For example, why would I assume that MIT is going to violate Massachusetts law and record audio? The cameras possessing microphone hardware is irrelevant if the software doesn’t turn on audio recording.

MIT clearly states in the FAQ linked in the article that audio isn’t recorded:

> No. The security cameras do not record audio. Massachusetts law prohibits recording audio without the knowledge of those being recorded. Additionally, we do not use facial recognition technology with our security cameras.

As you alluded to, 6-7 cameras per floor might be a lot or might not be. I would think it’s not that many: you’re probably getting two to cover stairwells, one for the elevator, one or two to cover a corridor. As soon as a corridor has a bend or runs a certain distance you’re getting another camera. We’re basically at 4 cameras per floor minimum for the smallest multi-story building possible.

How do you know that "we" in "we do not use facial recognition technology" covers ambient.ai, who presumably gets the feeds directly?

Nowadays since corps lie frequently I treat FAQs as another lie.

If the capability is there, it will be used, specially since the whole decision process appears to have been rather opaque.

> No. The security cameras do not record audio. Massachusetts law prohibits recording audio without the knowledge of those being recorded. Additionally, we do not use facial recognition technology with our security cameras.

could you read lips with AI?

> My high school also had cameras. Isn’t that rather typical?

Security cameras are incredibly typical. Having cameras that record _audio_ are not, and having cameras whose audio and video recordings are handed over to an AI company is _absolutely_ not typical.

Is there any equivalent to spraying a camera with paint for audio?

Might as well just smash them if you're already vandalizing. Spray painting is not a lesser offense here.

Many cameras include microphones that the user has no intention of enabling. These are like that, since they're in a state where no-consent audio recording is illegal.

No audio recordings are made or handed over to anyone. Handing video over to a service provider to do things like weapons detection is not uncommon and not inherently a privacy problem. It's done in hospitals today.

how accurately can a model do lip reading these days?

Based on the article I really didn’t feel like I could assume that they were really recording audio. I had a hard time believing that MIT was violating state law.

MIT very clearly states that audio isn’t being recorded:

> No. The security cameras do not record audio. Massachusetts law prohibits recording audio without the knowledge of those being recorded. Additionally, we do not use facial recognition technology with our security cameras.

Whether or not the cameras possess microphones is irrelevant.

Security cameras on the outside of a building to deter after hours crime is fairly normal.

Security cameras on the inside of a high school strikes me as deeply NOT NORMAL

Admittedly it's been many decades since I was at high school. I think some of your metaphorical frogs might be getting boiled here.

> Security cameras on the inside of a high school strikes me as deeply NOT NORMAL

1999 - https://sco.wikipedia.org/wiki/File:Columbine_Shooting_Secur...

2022 - https://en.wikipedia.org/wiki/Uvalde_school_shooting 2023 - https://en.wikipedia.org/wiki/2023_Nashville_school_shooting

I could go on ... but most of the school shootings in the us seem pretty well documented on cctv

Yeah as a non-American I can understand why it might be that way, but I think you should understand that globally this is not normal and is indicative of deep problems

If you're taking cameras in high schools because of school shootings as your baseline of normality, you probably need to calibrate a little differently

> in what universe is a handful of cameras (less than 10 per building) in a publicly accessible space on a private university unusual or alarming?

> My high school also had cameras. Isn’t that rather typical?

I don't think such a handful of cameras means the same thing today as it meant in e.g. the 00s. Lots of things have changed, the government, the politics, the culture, the technical possibilities.

> Lots of things have changed, the government, the politics

This reads like you're saying "When [Good Party] were in charge, they would have always done the right thing, but now [Bad Party] is in charge so we need to handicap the government (perhaps only temporarily, until we get in charge again).

More compelling would be "Doesn't matter if it's a Bernie Sanders Socialists supermajority or a Trump MAGA government, the government shouldn't be allowed to have such-and-such power."

No, I mean that before, the video would've led to a closet where maybe someone was occasionally paying attention, the footage just recorded in a loop, new overwriting old, and it would only be accessed upon request when an incident happened. The matter would be handled internally; people would use their heads.

Now, surveillance can be automated and centralized. Who knows where that footage is being taken. It's no longer necessarily the school handling its matters internally. They can just offload that and accept the judgment of an algorithm. Maybe you're holding a Doritos bag and that results in police swarming you and almost killing you for fear that it was a gun.

https://news.ycombinator.com/item?id=45684934

> Armed police swarm student after AI mistakes bag of Doritos for a weapon

Re: government and politics, maybe you fear your government is heading towards authoritarianism, something that would've never crossed your mind in the 00s, and that makes more conscious of surveillance. You've also become more conscious about Facebook and Palantir, etc, and how mass surveillance is a thing that's quite different from regular surveillance (handful of buildings, handled manually and autonomously).

It's not about what party. I don't think another Republican candidate would've had the same effect. Also W. Bush was on for most of the 00s; he was Republican, so what are you even talking about with parties? And though politics is a factor, I wouldn't say "When [Good Party] were in charge, they would have always done the right thing". Both ends of the spectrum can suck, and the polarity itself that's been growing in politics can be part of the problem.

Your high school has responcibilities over minors and the video never gets seen except when an incident is reported. Do you think office spaces or libraries need face tracking and ai cataloging surveillance too?

To me that is no different than having someone follow me and everyone else around and writing a report on all my daily activities and submitting it to arbitrary self-appointed authority.

Private office spaces can and do exactly that as long as it complies with the various privacy and biometrics laws out there. MIT is a private university.

I don’t say this as someone who is in favor of that sort of thing beyond a certain point.

Technically the university could literally send a real human to follow you around while you’re inside their buildings. Your professor actually kind of sort of does that while you’re in class and even bases your grades on your behavior, and can report code of conduct violations.

> My high school also had cameras. Isn’t that rather typical?

What? Absolutely not! At least not in the countries in which I went to school.

According to a quick kagi, over 90% of US high schools have security cameras.

> Look, I’m not trying to be pro-surveillance state or whatever, but

... but why do you then go on making all pro-surveillance state arguments ?

Is MIT a state?

When the data ends up inevitably in the state’s hands (almost certainly thanks to the AI company they’re contracting), it is 100% part of the surveillance state. Same reason telecom providers and Big Tech generally is part of the surveillance state; they’ll just hand over data, no questions asked, no warrants provided.

Let’s say I run some video cameras on my property, and I decide that I’m going to mail the contents of those hard drives to my police department every day.

That’s not the surveillance state. That’s just me as a private entity snitching.

MIT is not a public university. It’s private property not much different than my house. You’re generally not supposed to be in most of their buildings unless you’re a student, faculty, or staff, just like you’re not supposed to be at my house unless I invite you over.

I will concede, though, that some lines are blurred. E.g., MIT has a police force.

The other point of my original comment is really just that it seems like the most prestigious colleges have faculty and students that complain about stuff that’s really typical at less prestigious universities.

> That’s not the surveillance state.

Why not?

> That’s just me as a private entity snitching.

... to the state. You are surveilling for the state.

Domestic surveillance took off with McCarthyism.

What if the state threatens to take away all your money if you don't do that?

MIT has a police force, so in the way that matters for this particular topic, yes, they are a state.

Aside from that, the lines are very blurred these days, as the state often ends up gaining regular, easy access to systems such as this.

(Regardless, it's a bit in bad faith to bring up surveillance-state scenarios, and then, when called on it, claim that it doesn't matter because it technically doesn't apply to the organization at hand. Even though it does in the ways that matter.)

That’s not bad faith, it’s just basic logic. MIT isn’t a state and is just installing security cameras on its private property.

I do concede what you’re saying about its police force, and that’s a fair angle.

Sorry to interrupt the pedantry about whether a specific entity counts as a state, I'd like to submit the notion that maybe the argument shouldn't be "is a state entity" but rather "is an entity that has some form of authority over my livelihood that is cornering me into a mode of constant surveillance which I do not consent to."

The obvious answer is to leave the organization, but ultimately one shouldn't have to end their employment in order to be free of undue/unjust AI surveillance.

Yes, this. That's part of my assertion that this is an argument in bad-faith: focusing on irrelevant details in order to derail an argument.

Yes. They have borders, government, and a police force.

Man I wish HN commenters could be bothered to even attempt to read each other charitably instead of this crap

Not sure there's a more charitable read on "I don't want to be pro-surveillance, but [bunch of reasons why I think this is fine and that surveillance should be and has been normalized]".

whoosh

Yes there is. That's the point of the comment. They made their points seriously and people should engage with them seriously. Instead of this reductive tribalism: "oo you expressed doubt about whether this is really bad, you must be One Of Them". Well they actually doubt it. Now what? Does trying to shame them out of doubting it change their mind? Of course not. Either someone engages with their questions and convinces them to agree or they stay on the fence. Trying to shame people out of discussion by pointing out how they're acting like the Wrong People is anti-intellectual and disrespectful and ignorant. Not to mention completely ineffective.

Fair point; you're absolutely right.

However, I did follow up with a direct reply before making the comment you're replying to, so... I literally did "engage with their questions".

(to be clear, this thread is not in reply to your comment. There is some overlap though)

A charitable read of this article isn't even necessary to understand the point of this is the opposite of the poster you replied to. Basic discernment at most.

I did and then I asked a serious question.

These are arguments that are put forward in favor of mass-surveillance.

That is in direct contradiction to the claim "I’m not trying to be pro-surveillance state or whatever".

So I was wondering how that contradiction comes about and asked about it.

What would you suggest the charitable reading should be ?

You believe them. You believe the line "I'm not trying to be pro surveillance state", and then you take seriously the points they made which are completely reasonable things to ask, and you engage critically with them, and discuss how those points are compatible or incompatible with the overall stance. It is easy.

As a general rule: to engage charitably with someone you believe the things they say they believe, at face value and attempt to understand how they could come to believe those things.

But I do believe them. It's just that they points that they sincerely ( I did not doubt that) made are part of the same set of points that are made in favor of a surveillance state - there is a contradiction here. Instead of assuming, I asked a question.

Quite literally this:

> attempt to understand how they could come to believe those things.

Is it possible that you are not reading my comment charitably ?

You asked them "why do you go on making these pro surveillance state arguments?"

How is that supposed to be a serious question? They brought up real concerns. The "why" is obviously that they had real questions--that's the purpose of all questions. So they asked them,and then you implied that, if a person was truly not pro-surveillance, they would not ask such questions. Yet they had those questions. So you think they should have pretended to not be confused? Not asked those questions? Pretended like everything made sense? What was the goal here?

You didn't engage with the content of their questions at all. You just called out the act of questioning. You implied that by asking at all they were on the wrong 'side'.

Look, I was trying to understand why the op believed the points that they made were not "pro surveillance-state", that doesn't align with what you are interested in reading here, but that's not really my job.

> You implied

I think you are implying that I implied.

Best to end this here, probably.

Everyone who writes 'not trying to sound pro X, but... ' is about to follow that with things that sounds pro X. That's what that rhetorical device means. So they obviously knew that the points sounded pro-surveillance. That's why they wrote that disclaimer.

They sounded pro-surveillance because they are pro-surveillance. The better question should be why a pro-surveillance person making pro-surveillance points felt the need to start with "I'm not pro-surveillance but". Perhaps they know their position is very unpopular and are hoping this magic spell will ward off some criticism?

Or maybe it's just more complicated than being on one side or the other.

Engaging charitably with the bullshit means the bullshit wins.

https://en.wikipedia.org/wiki/Brandolini%27s_law

https://en.wikipedia.org/wiki/Sealioning

They don't, did you read it? It's clearly a satirical art project exposing the bullshit.

I mean the correlation is money. Rich people being outraged

did an AI surveillance startup write this? You gotta be kidding me