Apps like Temu or TokTok. Or those cheap electronic devices where you have to download a questionable app and register an account. What exactly is being stolen and what is being done with it? Who is doing it? Why?
Apps like Temu or TokTok. Or those cheap electronic devices where you have to download a questionable app and register an account. What exactly is being stolen and what is being done with it? Who is doing it? Why?
The biggest problem i have with my data being collected, analyzed and used is in the fact that it will almost certainly be used to teach a ML model about how to better manipulate with people like me - the people that are privacy conscious and are trying as much as possible to reduce their fingerprint.
That data is invaluable, and if there does exist a way how to target even people like that, which there probably does since we’re only humans after all, the ML model will eventually figure it out. And they have literally billions of people to experiment and learn on.
Now, we already know from a few leaked studies made by Facebook that they cab already pretty well manipulate people into mostly whatever they choose. Take a hypothetical situation where you get a crazy out-of-touch billionaire, who decides to buy a large social network company, and then decides “Hey, I really want this candidate to win. Tune up the algorithms!”.
And the ML models will get a clear goal, that has been already proven to just work pretty well at influencing user behavior. And any data you give them, it helps the model to fine tune into influencing people like you . Which would also be really hard to prove, because ML models are by definition black boxes that are really hard to reverse engineer, and proving that it was trained to do this is AFAIK almost impossible.
I don’t want no part in that. Thankfully, all the large social networks have CEOs that are reasonable and would never try something like that, right?
And one more thing - you may not think that data about your behavior are of interest to anyone right now. But look at China and their Social Credit. And imagine how would have I.e holocaust turned out, if the government had access to all the data, opinions and profiles of people that are being collected now.
Oh, you mentioned you sympathize with the Jews three years ago in a private message? Well, let’s hope the country you live in never ends up in a situation where that could be a huge problem for you or your family.
So, every time any site is offering a “personalized, curated list” for you (I.e the google search result, or YouTube recommended videos), assume you are potentionally being manipulated, and avoid the site altogether- because there’s no other way how to prevent it. The ML model knows that you know, and is already trying to figure out how to manipulate people that are taking care not to be. And if there is a way, it will figure it out with some success.
The potential future authoritarian government has been my primary concern when it comes to data collection and profiling by corporations like Google and Meta for years. The governments don’t even have to build their information gathering networks, although they still will, but so much of the surveillance has been done for them, goes back years (literally an entire lifetime for many people now), and is just a request away. I can’t judge how the climate will be in two years, let alone a decade or two from now, but that information isn’t going anywhere.
One of the projects I have in mind is to explore some kind of “offensive privacy”, where the focus would not be on not being trackable, but on your computer spewing random bullshit and behavior into the algorithm to confuse it, and have it learning on behavior that’s not really true, but only generated. This will enable you to kind of fight back and if done by enough users even reduce the effectivness of ML algorithms, since they would be learning bullshit. Unfortunately, the scale required to effectively affect the learning process of ML models would be enormous, so it’s not really feasible, but I think it’s still better than just “staying hidden”.
With the advances in AI, creating a tool like that, that would simulate several random user behaviors on your IP/fingerprint, shouldn’t really be that hard.
And as an added bonus - if it clicks on adverts, it’s costing someone money. Fuck corporations.
You talking about a more extreme version of something like AdNauseam?
Exactly! AdNauseam is my inspiration, unfortunately I dont use it right now because it doesnt work with my setup, due to Mullavad in combination with PiHole being too good at their job, so it cant really click on the adds even if it wanted to.
It would ideally require a solution that uses its own DNS server, so you can keep on using PiHole, and prefferably avoids living in your browser so you can keep it locked down for privacy.
This is straight out of the book Little Brother by Cory Doctorow. Love the idea.