I think Microsoft’s planned recall feature where they screenshot everything you do so that it can be analysed by AI isn’t as bad as everyone makes it sound. It’s only bad because Windows is closed source and nobody can verify if what they say is true.
But if Microsoft aren’t lying and none of the data ever leaves your PC (which is supported by the fact that you need a pretty beefy machine to use it) then it is one of the more privacy friendly shit they’ve done recently. And I think they were fully aware that they could only sell “thing that records everything you do” if they could convince people that it doesn’t share that data. Guess they failed.
If it were open source I might even think about using it myself. If the hardware and subsequently power requirements weren’t so absurdly high.
You must log in or # to comment.
The big thing here is consent. If you run it yourself, i.e. opt into it. Then it’s consensual.
Microsoft has demonstrated over a long period of time they are happy to force “optional” anti consumer things into people through
- Bad defaults
- Silent updates changing settings
- Nag screens
- More nag screens that pop up randomly hoping you misclick
- Deceitful UI (Yes! Ask me later!)
Oh! You have misunderstood the whole concept of privacy. I have a thought experiment for you:-
Let’s assume Microsoft is not lying 🤥. The data (screenshot) remains on device, which is passed to some AI model like Image-to-text etc. This model generates text on-device. But no where Microsoft guarantee’s that the text generated or output from those AI models won’t be sent to the Microsoft. They only say the screenshots and AI models remain on-device, but the output/metadata can be sent to Microsoft.
That is the issue. Earlier there were many apps where Microsoft couldn’t pry because they were encrypted etc. Now they don’t need to break any encryption they just need metadata. That’s easy to transfer and use.
And to take this one step further. Say you can trust Microsoft 1000%, they are on your team and do everything they can to protect you and your privacy. They are an American company and as such are subject to American laws.
So when elections happen somewhere in the future and some asshole gets elected, they can order Microsoft to use their systems and their data to figure out a lot about people. They can for example figure out who has been using a certain kind of software, who has been consuming a certain kind of content, who has been playing certain kinds of games etc. Then this data can be used to target specific people for punishment for example, a one way all expenses paid trip to a new series of gulag up in Alaska.
You need to be able to protect your privacy 100%, not just for your protection today, but also for the future.
Why would this asshole not be able to order Microsoft to start implementing this software after they got elected if they didn’t do that before he was elected?
Upvote for unpopular opinion.
This “feature” is like a cop following you or your vehicle 24x7. Sure, you aren’t planning on doing anything illegal. But do you really want a cop following you 24x7?
yeah which might be useful if you happen to get a flat but it aint worth it.
Governments can already subpoena chat companies (Facebook Messenger, Discord) for records. Imagine what they’ll get if they just have your desktop and they suspect you of wrong think.
Even worse, if they only get the completed model and then draw their conclusions from something the AI dreamt up.
But that’s why god invented full disk encryption.
Kind of a big “But” for a company that makes so much background stuff run with the same SYSTEM account and doesn’t even drop privileges to dedicated users like more sophisticated OSes (e.g. Linux) do.
Genuine unpopular opinion and the people downvote it… In a community named “Unpopular Opinion”.
An unpopular opinion can have more or less thought put into it and be genuinely interesting and get up- or downvoted accordingly. Just like a photograph in a photography sub can have more or less thought put into it and an interesting or boring subject and get up- or downvoted accordingly.
Genuine photograph and the people downvote it… In a community named “photography”.
Sounds like utter nonsense doesn’t it
Unpopular opinion communities will always have a problem, bc the opinions that get upvoted are the ones people think are unpopular but still to a certain degree agree with.
Honestly, I wouldn’t even know how to vote on something I don’t agree with. I just add everything together for my dopamine.
One small bug in implementation can make it a juicy attack vector. Sure you can configure it to give more privacy, but how many users will do such configuration? Many casual user didn’t even bother changing their user/computer name (how many times have you seen the user named “user”, “asus”, or any other default?)
- It’s completely useless, as you could just save everything you need, and almost all things can just be searched quicker using traditional ways. I have yet to hear one single use case for it.
- It will leak all your data. Even if it does not upload the pictures to M$’ cloud, it still saves passwords, messages and files you wanted to delete in clear text.
- The devices which will officially support the AI, called CoPilot+PC’s, will also have Microsoft’s “Pluton security chips”. Which is just another way to say “Hardware RAT”. Yes, it’s literally just a way for M$, police, government, etc. etc. as well as talented hackers to gain system level access to Microsoft’s (““your””) PC. They don’t need to hack every device individually, which you could protect against with brain.exe and up to date software, they just need to gain access to M$’ servers (which has happened countless times before) to use Microsoft’s built-in hardware trojan to just fuck everyone over in a few seconds. And what is Pluton supposed to do? Prevent you from using anything that is not allowed by your government or M$, be it illegal streaming services, Tor, other VPNs, Linux or just everything except state media. (North Korea loves this.) So in order to use this, you will not only store clear text logs of everything, from passwords over secret files to nudes and maybe even CP someone sent you to troll you, because the second you open it “Recall” will screenshot it, so even deleting the original file won’t help, analyze it and call the cops. Want to fuck over your neighbour you don’t like? Use some AI tool to create convincing looking stuff, and he’s gone. And even if nothing is leaked initially, using the AI literally requires giving every entity with enough “authority” or knowledge full access to your device and therefore secrets, files etc.
So not only is it nearly useless, it in itself, its implementation and requirements are basically the equivalent of a M$ controlled North Korean like dystopia. Except uninformed users won’t even know, until they need to pay 5k in bitcoin to get their data, cops show up at their door or their bank accounts are suddenly empty because a hacker gained access to M$ servers (again).
Found the Microsoft Simp.
Did you even read the post?
Never thought I’d see microscoff astroturfing on Lemmy. We’ve made it!
Thanks for the new words to filter on here 😘
You’re delusional if you really think MSFT gives a shit about this community. Call OP a Microsoft shill / simp, sure, but suggesting astroturfing is ludicrous
