Open any social app and the feed you see is built for one goal: keep you looking at it. That's not a conspiracy theory — it's the plainly stated business model. The question worth asking isn't whether your feed is curated. It obviously is. The real question is: curated by what, and how much does that actually shape what you end up believing?
Here's what we actually know, not just what people assume.
In 2015, researchers inside Facebook (Bakshy, Messing, and Adamic) published a large study in the journal Science looking at over 10 million U.S. users to measure how much the News Feed algorithm limited people's exposure to viewpoints they disagreed with, compared to how much people limited themselves through their own choices — who they friended, what they clicked.
The finding that made headlines: individual choices about what to click on had a bigger effect on narrowing what people saw than the algorithm did. In other words, you filtering yourself mattered more than the algorithm filtering you.
But that's not the whole story — and other researchers pushed back hard.
Critics pointed out real problems with how the study was framed. The comparison baseline mattered a lot: the algorithm did measurably reduce the diversity of what people saw, just less than people's own choices did. Comparing the two and declaring the algorithm "not the problem" struck several data scientists as a bit like being told a small amount of a harmful ingredient doesn't matter because a larger amount of a different harmful ingredient is already in the recipe. Both things can be true — you self-select into your own bubble, and the algorithm reinforces that bubble further — without either one canceling the other out.
Since then, follow-up research across different platforms has landed all over the place: some studies find real, measurable "filter bubble" effects; others find surprisingly little; a few find no significant effect at all when they test it directly by turning algorithms off for consenting users. The honest summary a decade later is: the effect is real for some people in some contexts, it's smaller than the popular narrative usually claims, and it interacts with your own choices in ways that are genuinely hard to untangle from the algorithm alone.
What's not up for debate, regardless of how you land on that research: every major platform ranks content by what's likely to keep you engaged, not by what's most accurate, most important, or most representative of what other people think. That's true whether you lean left, right, or don't think about politics online at all. The ranking system doesn't know or care about the truth of a claim — it knows what you're statistically likely to tap on next.
A few practical things that actually help, regardless of which side of the research debate turns out to be right:
- Deliberately follow at least one source outside your usual lane. Not to "win an argument" with it — just so your feed has something in it you didn't select for yourself.
- Notice when a feed feels too agreeable. If everything you're seeing confirms what you already think, that's worth a second look, whatever the cause.
- Go to the source occasionally instead of the feed. Check a topic on a search engine or a primary source directly, rather than relying entirely on what happened to be ranked highest for you today.
Your feed was never a neutral window onto the world. It never claimed to be. The useful move isn't outrage at that fact — it's just remembering it's true every time you open the app.
Solitaire FactStack drops one media literacy fact into every game of Klondike solitaire you play — free, ad-supported, and built to make you a little sharper one hand at a time. Play free now →