I can't remember how exactly, but I came across adsettings.google.com, and it was sorta bitter-sweet. I liked that I could see (some of) what they've assumed about me.

Ignore the title, it's ultimately irrelevant. I just want to chronicle some stuff I learned; about activity, data and ads.

These components interact on systems. These systems operate intelligently, narrow and low, but still intelligent, where the general directive is to retain attention. They do that by asking for information or inferring from other indicators.

An experiment

Take Google's systems. From activity, system-wide, they'd make assumptions like: "likes anime", or "has children".

The experiment was to, over a period of a month, deliberately do things unnatural to me, as well as reject accurate suggestions as wrong, in a bid to change some of these assumptions they've made about me.

Even with closer inspection, one can't say for sure what activity directly impacted a result. Nevertheless, I reason directly telling them the wrong things ought to carry some significant weight, and if they get that wrong, they can go on to get others wrong. Two steps.

One, I searched for and browsed information, I assumed parents would search for. And after a while, in the suggested news articles came pieces about parenting. The indicator of "parent" also changed.

Second was marking accurate suggestions of videos to watch as "not relevant", watching videos I really had no interest in, as well as searching for them. Of course, video suggestions got weirder, to the extent that I'd open YouTube, scroll a bit and close it without watching anything.

I assume that performing similar steps on other ad-powered platforms like that will yield similar results.

It would take a lot of effort to not yield to instinct but depending on what you're looking for, it'd be worth a bit more. Compared to "deleting" your account and staying off the service.

Patient misinformation.