You always had access to see what your friends were watching on your own server. This is a consequences of being an admin, you kind of have to have access to that kind of data to manage your system and streams.
This seems to just extend it to showing you what they’re watching on other servers, as well.
Anyway, if the concern is that Plex, the company, has access to this data, then yeah, you probably should have read the privacy policy a little closer.
Jellyfin is there and doesn’t have a parent company to “phone home” data to.
It’s unfortunate that Jellyfin is just slightly worse than Plex at pretty much everything. Playback is smooth, sure, but set up is harder, getting good metadata is harder, logging in is harder, etc.
The metadata one really put me off. I set up a Jellyfin instance with the exact same media set as my Plex instance, and it immediately started “recognizing” standard movies and shows as porn and hentai. I’m still going to push through and get it properly set up eventually, but even so, I’m not looking forward to manually managing accounts when people can just SSO with Plex.
I mean, I have a ton of media that Plex recognizes automatically and Jellyfin doesn’t, so… Agree to disagree, I guess. I’m not trying to defend Plex’s recent enshittification, but that doesn’t change the fact that it’s generally a better experience than Jellyfin right now.
Anyway, if the concern is that Plex, the company, has access to this data, then yeah, you probably should have read the privacy policy a little closer.
come on, you know this is a non answer. also plex shouldn’t have this data, it should be for the admin only.
What? Plex is not one of those open source, self-hosted, privacy-centric services. Plex can do whatever the hell Plex wants with your watch history, because you agreed to their broad terms of service that said exactly that when you signed up. You chose to run your traffic and authentication through Plex servers because it’s convenient, not for privacy reasons.
If you don’t like it, use Jellyfin. I’m personally looking into moving, as Plex seems to be getting slowly shittier.
why are you defending them? sure, they’re allowed because they’re a big company so they make the rules, but that doesn’t mean you have to lick their boots and say oh actually that’s fine you made the choice. even big companies SHOULD be ethical. we DESERVE ethical treatment, furthermore, even people who didn’t wade through the terms.
I don’t know how you could read that and think I’m defending them.
I’m just telling you how the world works. If you want real privacy, you need to PAY somebody with a rock-solid privacy agreement or fully host it yourself. Plex is neither of those things. Remember, if something that costs money to run is free, then YOU are the product.
They say they use it to sync up your watch history to your account so it can sync across devices, but I wouldn’t be surprised if they were selling your watch telemetry to advertisers as well.
It’s a non-answer that their privacy policy explicitly states that they will collect this type of information and that they stipulate what kind parties they can share that info with?
That’s the straightest answer that you’re going to get. Privacy policies like this are bullshit, but they’re also the norm so acting like it’s a non-answer after 20 years of this being the norm seems a little… naive, perhaps?
he wrote “On my way to blow up the plane (I’m a member of the Taliban).” in a private group chat on snap chat
…a private group chat. Nothing stupid like posting it on xitter or other public place.
Its a fucking in-joke. Do I need to worry about what I say to my friends now in private and worry about what my friendly local government spy would think about it… ?
All this invasion of privacy all these years and all they have to show for it are a few false positives.
All the airport wifi could do is see the DNS requests (and the modern trend is to have DoH or DoT enabled by default, for example in the up to date versions of Android)
A court in Madrid heard it was assumed the message triggered alarm bells after being picked up via Gatwick’s Wi-Fi network.
Public wifi without a VPN is like sex without a condom. The connection may not be encrypted (very risky) and even if it is, you are still susceptible to man-in-the-middle attacks: garlandtechnology.com/…/how-to-monitor-encrypted-…
I guarantee there will be a flood of articles about this over the next few days because of what I quoted above.
It’s also possible that one of his “friends” reported him or something like that.
“A key question in the case was how the message got out, considering Snapchat is an encrypted app.
One theory, raised in the trial, was that it could have been intercepted via Gatwick’s Wi-Fi network. But a spokesperson for the airport told BBC News that its network “does not have that capability”.
In the judge’s resolution, cited by the Europa Press news agency, it was said that the message, “for unknown reasons, was captured by the security mechanisms of England when the plane was flying over French airspace”.”
Please explain to me how using Public WiFi is unsafe if the traffic is encrypted with TLS. Unless they somehow installed a keylogger on everyone connected to said Wifi and picked it up from there, the only way this was possible was on some quick text analysis and recognising the IP address from Snapchat
On its website, in a section titled “How We Work with Law Enforcement Authorities”, Snapchat says one of its goals is to “maintain a safe and fun environment where Snapchatters are free to express themselves and stay in touch with their real friends”.
It adds: "We also work to proactively escalate to law enforcement any content appearing to involve imminent threats to life, such as school shooting threats, bomb threats and missing persons cases, and respond to law enforcement’s emergency requests for disclosure of data when law enforcement is handling a case involving an imminent threat to life.
Yes, especially in the UK, since they’re a surveillance state.
There are some things that will always get flagged on any platform. This, drugs, and connections to sanctioned countries, for example. I’ve heard of people in the US having their Venmo accounts suspended because they put “Havana” in the transaction description. Havana is a local dance club.
Obviously this is a privacy community, and this ain’t great in that regard, but as someone who’s interested in AI this is absolutely fascinating. I’m now starting to wonder whether the model could theoretically encode the entire dataset in its weights. Surely some compression and generalization is taking place, otherwise it couldn’t generate all the amazing responses it does give to novel inputs, but apparently it can also just recite long chunks of the dataset. And also why would these specific inputs trigger such a response. Maybe there are issues in the training data (or process) that cause it to do this. Or maybe this is just a fundamental flaw of the model architecture? And maybe it’s even an expected thing. After all, we as humans also have the ability to recite pieces of “training data” if we seem them interesting enough.
They mentioned this was patched in chatgpt but also exists in llama. Since llama 1 is open source and still widely available, I’d bet someone could do the research to back into the weights.
Yup, with 50B parameters or whatever it is these days there is a lot of room for encoding latent linguistic space where it starts to just look like attention-based compression. Which is itself an incredibly fascinating premise. Universal Approximation Theorem, via dynamic, contextual manifold quantization. Absolutely bonkers, but it also feels so obvious.
In a way it makes perfect sense. Human cognition is clearly doing more than just storing and recalling information. “Memory” is imperfect, as if it is sampling some latent space, and then reconstructing some approximate perception. LLMs genuinely seem to be doing something similar.
I bet these are instances of over training where the data has been input too many times and the phrases stick.
Models can do some really obscure behavior after overtraining. Like I have one model that has been heavily trained on some roleplaying scenarios that will full on convince the user there is an entire hidden system context with amazing persistence of bot names and story line props. It can totally override system context in very unusual ways too.
I’ve seen models that almost always error into The Great Gatsby too.
This is not the case in language models. While computer vision models train over multiple epochs, sometimes in the hundreds or so (an epoch being one pass over all training samples), a language model is often trained on just one epoch, or in some instances up to 2-5 epochs. Seeing so many tokens so few times is quite impressive actually. Language models are great learners and some studies show that language models are in fact compression algorithms which are scaled to the extreme so in that regard it might not be that impressive after all.
How many times do you think the same data appears after a model has as many datasets as OpenAI is using now? Even unintentionally, there will be some inevitable overlap. I expect something like data related to OpenAI researchers to reoccur many times. If nothing else, overlap in redundancy found in foreign languages could cause overtraining. Most data is likely machine curated at best.
Honestly the thing that annoys me the most about this isn’t the privacy aspect. It’s the fact that they called it “new outlook”. Which means now at work I have to explain that no, this isn’t real outlook it’s just MS being useless wankers and not being able to come up with a new name for a new product. See also, teams vs teams for work and school. They did the same thing with Skype and Skype for business back in the day and still pisses me off.
This is such a weird thing I’ve noticed on this community. There was a guy not too long ago that would make new accounts like daily so he wasn’t posting under the same username and it’s like… why?
I get you want privacy, but there’s a line where it just stops making sense, and your personal info isn’t that valuable. Anyway
I get you want privacy, but there’s a line where it just stops making sense, and your personal info isn’t that valuable. Anyway
Actually, you don’t need perfect privacy. You just need good enough privacy, and here’s why:
If you’re a low-value target - i.e. a random internet user, that’s you and me - always remember that your value is low: Google, Microsoft, Amazon, Facebook… expend a certain amount of resources to fish for enough of your data to earn them a return on their investment. We’re low-value targets, so they first and foremost go for the low hanging fruits: the people who don’t know, don’t care, wallow in social media without any restraint and make it particularly easy to gather data from.
All you have to do is make it hard enough and expensive enough for the corporate surveillance collective to lose money on you: create accounts full of fake data and don’t post personal information - or make up fake personal information in your posts - to poison their wells. Don’t post photos of you or your family. Use throwaway email addresses. Use a deGoogled phone. Don’t browse without an ad blocker set on reasonably high. Use a browser with anti-fingerprinting. Don’t fill out Costco membership cards. Pay with cash stuff that you don’t want anybody to know about. Etc etc.
In other words, adopt a reasonable-enough privacy hygiene so that you’re not part of the low hanging fruits. It doesn’t have to be drastic, just good enough to make you not worth the sonsabitches’ time and effort.
If you’re a high-value target however, a Snowden or an Assange, that’s a different proposition. But for the rest of us, private enough is good enough.
That’s why places like Lemmy and Mastodon are nice, even if big corpo buys up some instances, there’s still the option to just start free ones elsewhere.
In many unzip utilities, they use temp files that you wouldn’t be paying attention to. These temp files will contain your credentials and you won’t know where they are or if they got deleted.
And even if they’re deleted by the archive program, it’s likely a normal deletion, and not a secure delete where the original data is overwritten with random data before deleting the entry in the file system, which could be potentially recovered.
And yet thousands of security researchers can’t find a shed of evidence. This shit is tiresome and counter productive. The general public is weary of hearing this made up bullshit.
The technical practice isn’t hard. That’s the claim. The reality is nobody is buying shit doing this and this is just another repost from the same 404 article months ago.
Bro, I’ll literally be having a conversation with someone about a topic, and all of the sudden Google starts recommending me products related to the discussion afterwards. Smart phones and smart speakers without a doubt listen in on our conversations. There’s the evidence.
Eh, surprised that’s happening to someone in this community. Strip Google off your phone and throw out any hardware with a microphone that doesn’t run open source software and this will stop happening.
privacy
Top
This magazine is from a federated server and may be incomplete. Browse more on the original instance.