Do you think tech companies know you better than you know yourself?

[noname223]

Pro: I am a mass user in an online forum and posted a lot.
I spend a lot of time online and I use mainstream services
I think there are a lot of information tied to my accounts even if they cannot connect all to my real identity
I think machine learning could make impressive predictions about my (online) behavior and my current mental state
I feed the algorithms with using Youtube and search engines
Sometimes I am unaware of things when I am paranoid a machine might spot that or in case of chatGPT make it worse and lead me into an AI psychosis


Con: The information is very selective and I think I only share some types of information there are some gaps
I think when I use AI I notice it gets a lot of things wrong. And sometimes it makes false inferences and just guesses something

The answer is a yes and no. It depends.

 

[Dejected 55]

No. It is contextless unqualified information that the bots gather about us. The closest it gets to "knowing" us are the places we shop that track our purchases and try to recommend other things to us... but even they they are recommending not based on us but trying to find connections between our patterns and other people's patterns... and if you don't fit the pattern, the suggestions will make no sense.

 

[ObsidianMidnightSky]

No. I turn off cross-site tracking, ad tracking, personalized recommendations, learning from behavior, etc.

I use temporary and non-learning AI for most things so an accurate profile cannot be made.

My data profiles mostly consist of the history I allow and very general location. I often manually delete history and only keep the things I actually want to influence the algorithm.

Unfortunately, this creates a very annoying experience of unrelated and offensive topics I have no interest in consistently popping up in advertisements and less-than-accurate recommendations, but I do prefer data-sellers to have mangled or limited information.

Note: One of the fears with TikTok was its ability to profile enemy countries to feed targeted information to socially control/win wars. Big tech profiles can certainly be weaponized, but you wouldn't even know it. Data correlations can certainly be made for things you wouldn't have put together, such as predicting your mental health state and vulnerability before you admit it to yourself. This can be used for good, like in healthcare systems, or bad, like in corrupt governments.