A place to share alternatives to popular online services that can be self-hosted without giving up privacy or locking you into a service you don't control.
Rules:
Be civil: we're here to support and learn from one another. Insults won't be tolerated. Flame wars are frowned upon.
No spam.
Posts here are to be centered around self-hosting. Please ensure it is clear in your post how it relates to self-hosting.
Don't duplicate the full text of your blog or git here. Just post the link for folks to click.
Submission headline should match the article title.
No trolling.
Resources:
Any issues on the community? Report it using the report flag.
Questions? DM the mods!
One of the best pieces of self-hosted software ever to exist.
Edit: This is Immich! for the folks who don't know.
My read is that FUTO as a software movement is totally fine, it does what it claims on the tin. The people behind FUTO are a different story, and the main person bankrolling it seems to have friends with odd views (I think they're blown out of proportion, but they're still concerning).
You'll never find a perfect movement. Here's what FUTO seems to prioritize:
That sounds pretty good to me! I'd prefer it to be FOSS, but allowing me to distribute modifications for non-commercial use is probably good enough for most things.
I probably disagree with their founder politically, and I'd run FUTO differently, but I think their software is good and I could maintain it myself if needed, and at the end of the day, that's what matters to me.
FUTO doesn't seem interested in getting involved in politics, they're merely musing philosophically, and their products aren't profitable, so it doesn't really matter to me what their political positions are.
Software can't be fascist, it's just software. The makers or users can be fascist though. If that statement was true, Lemmy would be tankie.
No, that's not fascist. Facial recognition software can be used for a variety of reasons, like unlocking a phone or laptop, gaining access to secure areas, or home automation stuff.
It's only fascist if used by a government to oppress minorities. The software itself cannot be fascist, but it can be used by fascists.
The fault lies with the makers and users of the softeware. Software doesn't have political opinions, it's software.
It's like saying Panzer tanks were fascist because they were built by the Nazis. Tanks cannot be fascist, they're tanks. So despite being made and used by fascists, they're not fascist, they're tanks.
That's the same exact thing here. Facial recognition software can be used by fascists, but that doesn't make the software itself fascist.
The other person deleted their comment so I can't really know what the argument was, but I would like to make a distinction:
While tools cannot be political themselves, tools can lend themselves to specific political purposes.
A tank cannot itself be fascist, but it can make fascism more viable. Surveillance software cannot be political, but it is easily abused by fascists to destroy political opposition.
What matters is the harm and benefits. Is the harm caused by the tool justified by it's benefits? Or are the primary use cases for the tool to prop up fascism?
(I suspect that "authoritarianism" would be a better term to use here, but I'm continuing the theme of the thread)
Their argument was that software can, in itself, be fascist, and that's what we went around and around on. The example given was facial recognition software that can determine race (and later, country of origin).
Essentially, I said exactly what you're saying, while they argued the opposite. I wish I quoted them, but I did only directly address their claims, if you'll take my word for it.
I don't want the government to have and use facial recognition software (their example) and extensive security camera systems (my example, such as Flock), not because those solutions are fascist in and of themselves, but that they can be used by fascists to accomplish their goals. Even if the current regime uses them purely for good (i.e. completely opt in facial recognition, cameras inaccessible to police until there's a warrant with no passive collection) the next regime may not.
The extension of the argument I'm making (and maybe them kinda?) is that it's functionally the same as if the software were political.
You can make software that nearly exclusively benefits a particular political belief for family of beliefs.
So even if it's not actually technically political, it can be functionally political, at which point the argument is splitting hairs.
I think those are important hairs to split.
Let's say there's a camera system built due to a direct public vote and rolled out by a political party all agree defends democracy. The stated goal is catching red light violations and speeders, and it's a popular system. As part of the functionality it reads license plates, and that is verified by a human every time, and no footage is stored if there's no violation.
Is that system fascist? Most would say no, and it exists in many states, like California and Washington.
Then the next election, a fascist is elected, and one of the first moves is to repurpose that system to track undesirables, and now it stores a ton of footage.
Is that system now fascist? It's the same exact system as in the previous example, it's just being used for fascist ends, such as tracking vehicles with certain plates (e.g. Illegal immigrants, minorities, etc) Nothing has changed in the capabilities or programming of the system, the only change was when to capture footage, what people use it for, and how long to store it.
Yes, it's theoretically possible to design a fascist system, such as an LLM that only gives fascist answers, but that's an incredibly narrow definition.
Just because a product has a plausibly deniable use case doesn't really mean that it's not functionally political.
If someone creates a super invasive surveillance system and initially uses it for a seemingly benign purpose, that doesn't mean the intention all along wasn't more nefarious, especially if the system was practically irresistible for power structures and it's use directly lead to authoritarianism. Like giving someone their first hit for free.
In a case like that, I would discount the benign use as a red herring, and say that the software is functionally political.
The intention can be fascist, sure, but that doesn't mean the solution is fascist.
For example, I think it's pretty clear that Lemmy was designed by tankies to create a safe space for tankies (why would the instances the main devs maintain be overly protective of China and Russia if it weren't?), but that doesn't make Lemmy "tankie," it's a software project that can be used by fascists, tankies, commies, anarchists, statists, etc, because it's just a software program.
Likewise, a surveillance system can be used by a fascist government, private company to protect company secrets, government agency like the Pentagon for internal use, or even private individuals to ID who is at the door. It's only fascist of it's used to further fascist goals, like identifying minorities or protestors. But then, it's still not the software that's fascist, but the whole system, meaning how people use it and the policies in place.
The chance of a given piece of software being "fascist" is incredibly low, since it would need to act in a fascist way and only a fascist way, or only be useful for fascist ends. Like the fascist LLM example I gave, or a training simulator that is hard-coded to only present fascist ideology.
Right. That's what we're talking about.
But I think the bar is a little lower. I think it's enough to be primarily useful for (eg) fascist goals. If it happens to have minor non-fascist uses, I don't think that materially changes anything.
I don't think that Lemmy is primarily useful for furthering tankie goals.
I think that privacy invading surveillance systems are primarily useful for furthering authoritarian goals, by intention or not. There are some nice alternative uses, but I think that the use case of primary importance is in service to authoritarianism, which makes it authoritarian software.
And I disagree. I think this all started when we allowed things like traffic light cameras, speed cameras, and toll cameras to automatically bill based on license plate. I don't think most would consider those to be "primarily useful for furthering authoritarian goals," they're merely there for routine law enforcement with specific goals.
Flock cameras are basically that same exact system, but instead of only being used when something tangible is triggered (red light, radar, or toll booth motion sensor), they passively collect information. Flock is a private company that sells its surveillance services to cities (and private orgs) to assist with tracking down license plates or alerting when there's a gunshot detection. This is allegedly legal because you don't have any expectation of privacy when you're in public (hence why Ring doorbells are legal), and private companies don't have to follow the same rules as law enforcement. I personally don't think Flock's founders are fascist, they seem to genuinely want to help reduce crime. I worked for a similar company that mostly did perimeter security (i.e. generally only operated on private property), and the founder was absolutely not fascist, but they did want to help reduce crime.
I personally don't consider either of those systems fascist by nature, but they can be used to achieve fascist goals. Tracking burglars across neighborhoods doesn't sound especially fascist to me, but tracking protestors certainly does. These are very dangerous technologies that can easily be used for fascist purposes, so I think we shouldn't allow them to be used at all, not because they are fascist, but because they can easily be used for fascist ends just by changing conventions around its use.
I don't think we need to label a system as authoritarian or fascist to oppose them, we just need to point out how easily they can be misused.
So this is why I'm trying to avoid using the term fascist, because it means something specific but nobody can really agree what that thing is. For the purposes of this discussion, I'd prefer to say "authoritarian".
I wouldn't call traffic cameras invasive because they're only at (some) intersections. But it's still kind of borderline.
A private citizen recording people in public and the government doing so are fundamentally different. I think that having the government subcontract away that responsibility to maintain privacy is an abdication of that responsibility and is an intentional act to move towards authoritarian on the part of the govt. Now if the private company intends to help the government do that, is immaterial; that is the only major use case for their product, so it is functionally a tool with an authoritarian purpose.
Is it such a dichotomy in reality? No.
But we need to be exceptionally careful when we see these gray areas
It's more that people probably know what it means, but choose to misuse it to smear their political enemies, and then other people who don't know what it means repeat it.
Here's a clear definition in case you or anyone else that reads this isn't clear on it (or pick your favorite dictionary, it'll be similar):
Is a network of cameras with facial recognition fascist according to that definition? No. Is it useful to people pushing for such a government? Yes. Is it useful to other authoritarian systems of government? Yes. Is it useful to non-authoritarian systems of government and non-government entities, including private citizens? Also yes.
What if they're at every intersection, stop signs included?
If the only thing that turns something into an authoritarian system is scale, then it's not the system that's authoritarian, but the way they're used that is authoritarian.
I oppose red light cameras not because they're authoritarian in and of themselves, but because they can be used by authoritarians to screw people. I oppose Ring doorbells not because they're authoritarian, but because the corporation has control and can hand that data over to authoritarians without consent from the owner (or be compelled by authoritarians).
"Authoritarian" is an adjective that describes people, governments, or policies, not inanimate objects or software systems.
Exactly! The capability to record the public isn't authoritarian, the government policy of recording the public is authoritarian.
This may sound like a pedantic point, but I think it's an important one. If cameras are authoritarian, then ban cameras and the problem goes away right? The government will just use radar, track financial transactions, or something else entirely, and you have the same problem.
The real problem isn't cameras or facial recognition, but that the government tracks people. To solve that problem, we shouldn't ban the various ways the government can track people, we should ban the government from tracking people. Don't b regulate the tools, regulate the people using the tools.
I mostly agree with you, so we're probably not really doing much in this discussion. I'm trying not to be pedantic, but as my name will tell you, I find that to be a challenge lol.
I agree wrt how to regulate.
If disallow the govt from broad indiscriminate surveillance and disallow the govt from circumventing that rule by subcontracting it to private entities, then these companies and products that perform the mass surveillance would naturally become unprofitable and collapse. I would argue that such a product would be by its nature political, because it's only practical use case was the furtherance of a political goal.
Cameras aren't political, but the use of cameras for mass govt-level surveillance is political. So a system that does so (like the ones sold to the govt) is a political software product.
To me where it gets tricky is when private entities grow to government-sized proportions, and begin to use these same tools for similar purposes. I think that is also a problem, but it becomes harder to frame it.
Again, I disagree. Surveillance has a lot of use cases outside of government, and a huge use case is keeping the government in check. Palantir could have sold its services to non-profits like the ACLU as a check on local, state, and law enforcement agencies. They could have sold it to HOAs and neighborhood watch associations as an early warning system for repeat offenders.
The government skirting the 4th amendment (and a few others) doesn't automatically make its sub-contractor's products "authoritarian," it makes its use of those products authoritarian.
I disagree with that conclusion. The use by the government is authoritarian, but that doesn't make the product authoritarian.
A private entity can do authoritarian things, like spying on its employees or customers. Authoritarianism isn't strictly tied to governments, but anything that acts like a government. Here's the first definition I found:
Software can't really favor obedience to authority, it can't really deny you your freedoms, it's just software. Likewise for a camera system. The only way those things can be authoritarian is if paired with some form of enforcement arm, like corporate security or law enforcement. So that combined system is authoritarian, the cameras or software on their own cannot be authoritarian.
That's my point.
In theory, yes. In practice no.
ALCU could not roll a system like that out; never mind securing the resources needed to deploy this meaningfully; using it would go against their ethos, because using it would make them authoritarian, or adjacent.
Similarly, even if HOAs could deploy a system like that, that'd make them authoritarian.
Mass surveillance products like these don't have a lot of non-authoritarian uses. Even if you could find such a use (of which I'm skeptical), it'd almost certainly need to be subsidized by an authoritarian customer. We're not talking about security cameras around you personal property, here.
That really depends how the system is used. If it explicitly doesn't record regular residents and people who have signed up officially as visitors (and homeowners can review footage), I don't think the camera system itself would really be authoritarian. Yeah, the system would be capable of violating privacy, but as long as the system is transparent and reviewable by the residents, I think it can be privacy-respecting. Basically, it would be like a home security system, but across a neighborhood, and it can even be self-hosted to not let third parties access the data (and police requests would go through the HOA board, which consists of residents).
That's my point. If the system itself can be used in a privacy-respecting way (and the vast majority can), even if it's typically not used that way, the system itself cannot be authoritarian. If an institution uses it in an authoritarian way, then the institution is authoritarian.
In short:
I have friends that use home cameras to do object classification as a hobby, mostly to identify and fee record wildlife. I've also heard of people doing this to identify package deliveries and catch package thiefs. Sharing those models with others on the internet is largely the same idea as what flock is doing, and with enough data, similar solutions to what Palantir is doing could be done entirely by hobbyists.
The products Flock and Palantir aren't authoritarian in and of themselves, it becomes authoritarian when those products are used to enforce policy.
Again, in theory in a vacuum, I agree. But I disagree that anything you describe could actually be both commercially viable and deployable without authoritarian involvement
In your example do you not see all the gymnastics and bending over backwards you need to do to avoid the inherent nature of the system? I'd go so far as to say that the people in your theoretical HOA are analogous to supporters of a authoritarian regime.
You're making a pro-gun argument here, and it's not convincing for similar reasons: products are more than the sum of their parts, and the actual application of a product matters more than the theoretical use. If it is nearly impossible to meaningfully use apolitically, then it is not apolitical.
You haven't heard of Ring cameras? Commercial security systems? They do basically what I'm describing, just not as well because they don't have as much of an incentive. Are end users willing to pay for these more advanced models? No, so consumer grade cameras stick to object detection like deer vs racoon instead of specific individual detection (e.g. scanning eyes).
Governments, however, are willing to pay that amount. Why? Because they think it'll help them detect criminals, and they think that helps keep people safe. It's an extension of the HOA idea, just with government-scale funds backed up with law enforcement to go after threats. That, in itself, isn't authoritarian, but setting up such a system opens the door for authoritarians to take control and misuse it.
Analogous, sure, but the HOA has no enforcement arm for non-residents, so all they can do is ask the police to intervene. That's the difference with a city, it has a police force it can order to intervene using information from that system. It's the mixing of enforcement and surveillance that makes it authoritarian.
So a surveillance system is not itself authoritarian, it's only authoritarian of there's some enforcement arm to enforce obedience or punish disobedience.
Again, I disagree. Something is only political when used for political ends.
Well there you go. It could be authoritarian, except an authoritarian govt isn't subsidizing it. Exactly like I described.
You keep walking straight into the points I'm making.
Wrong. Setting up a super invasive surveillance system is inherently authoritarian, even if they initially happen to use it for reasons that don't typify authoritarianism. You have to bend over backwards so hard to keep it from becoming authoritarian, that it will just naturally corrupt any entity that deploys it, even making the monumental assumption that an entity that deploys this didn't have the intention to use it for nefarious purposes from the start.
Is a rather clumsy piece of mental gymnastics. Not only have you said it before. You can use this argument, coupled with your earlier "it's constituent parts aren't authoritarian" to argue that nothing is authoritarian.
And again this is just the pro-gun argument. Fine on paper, useless in reality.
I'm making the argument that it is possible for software to be political even if it wasn't created as such. I only need to show that a single case is possible.
You are making the argument that it is impossible, and you keep trying to prove it by example.
Again, it's not the software itself that's fascist, it's what it's being used for that's fascist. Facial recognition for determining citizenship could absolutely be used for non-fascist purposes, like simplifying border crossings to not require documentation (i.e. completely opt-in). Likewise, surveillance systems can also not be used until there's an actual warrant (i.e. no passive recording), which can help in catching dangerous criminals.
The technology itself isn't fascist, it's how it's applied that's fascist. The mass data collection is fascist, the tools used to collect that data isn't fascist in the same way that guns and tanks aren't fascist, but they can certainly be used by fascists.
If anyone is refusing to engage, it's you. You provided no argument for your position, whereas I've explained as best I can in detail, with examples of similar things. Me not agreeing with you isn't "refusing to engage," it's a good faith debate.
If there's some point you've made that I've failed to address, I apologize, I tried to be thorough to not waste any time going back and forth.
Hahah you mean like Lemmy itself?