prism
← Back to blog

The Difference Between Social Media and the Algorithm

Isaac Vesterfelt

In the last post, we looked at why deleting social media doesn’t work very well, and why “just quit” advice tends to ignore the fact that these platforms are infrastructure, rather than just entertainment.

This assertion raises an obvious question. If quitting isn’t the answer, and restriction only helps a little, what is the answer?

Here’s the thesis of this entire series: social media and the algorithm are not the same thing, and most of the harm comes from one, not the other.

This distinction sounds obvious once you say it out loud; however, almost none of the public conversation, or the apps marketed to “help,” actually treats it as two separate problems. They get bundled together as one undifferentiated “screen time” issue. That bundling is the mistake.

Two very different things wearing the same outfit

When you open Instagram, you’re actually doing two categorically different things, often within the same five minutes:

Thing one: You check messages from friends, see a photo your sister posted, respond to a comment, coordinate plans in a group chat.

Thing two: You open the Explore page, or get pulled into Reels, and an algorithm you’ve never seen and can’t inspect starts feeding you an endless stream of content from accounts you don’t follow, optimized to keep you watching.

The first is communication. The second is a recommendation engine. They happen to be bolted together inside the same app, with the same blue icon, but they are not the same activity, and they don’t affect you the same way.

The research backs this claim up

A growing body of research distinguishes between active social media use, messaging, commenting, posting to people you know, and passive use, scrolling through content without interacting. The pattern that shows up again and again: active, relational use tends to support feelings of connection, while passive consumption of algorithmic content is more consistently linked to lower well-being and a weaker sense of social belonging.

This association corroborates with something we covered in the first post: Harvard researchers found that routine social media use (checking in, responding to friends, treating it as part of your normal day) was associated with better mental health outcomes, while emotionally driven use (compulsive checking, FOMO, doomscrolling) was associated with worse ones. The “good” use case and the “bad” use case were happening on the same platform. The difference was what kind of use it was.

What the algorithm actually optimizes for

It’s worth being precise about what “the algorithm” means here, because it’s not a vague boogeyman. Recommendation systems on most major platforms are built to maximize engagement: the likelihood that you’ll click, watch, comment, or share. Researchers at the Knight-Georgetown Institute describe the core problem clearly: platforms have incentives to design these systems so that the items most likely to capture attention get ranked to the top, regardless of whether that aligns with what you’d actually choose if you stepped back and thought about it.

That gap, between what grabs your attention in the moment and what you’d actually want for yourself, is the entire reason infinite scroll and autoplay exist. The American Psychological Association has specifically flagged infinite scroll as risky, particularly for younger users, because it removes the natural stopping points that would otherwise prompt you to put the phone down. There’s no “end of the page.” There’s no moment where the app says “that’s all for now.” It’s designed that way on purpose.

So what happens if you remove the algorithm but keep the social part?

This is the part of the story where it gets interesting, and where the research gets appropriately humble.

A large-scale study published in Science took Facebook and Instagram users and switched their feeds from the normal algorithmic ranking to a simple chronological feed, just showing posts in the order they were posted, no recommendation engine involved. The results were a mix of good and complicated:

The honest takeaway here isn’t “algorithms are evil and chronological feeds fix everything.” It’s that algorithms are doing several things at once, some of which are clearly bad (maximizing time-on-app regardless of your goals) and some of which are more complicated (content moderation, filtering low-quality material). Simply removing the algorithm and replacing it with nothing isn’t automatically a win. The Georgetown researchers make a similar point: mandating chronological feeds is “simple to understand” but doesn’t necessarily address the underlying issue, which is that these systems aren’t designed around what’s actually good for the people using them.

The real distinction isn’t “social media vs. no social media”

It’s: content from people you actually know and chose to connect with, versus content from an algorithm selected for you from strangers because it’s statistically likely to keep you scrolling.

The first category is genuinely useful. It’s how people coordinate, stay close to family during deployments or moves, get support during hard times, and maintain the kind of “weak tie” relationships that, as Pew Research has found for over a decade, are one of the main reasons people use these platforms in the first place. Two-thirds of social media users say staying in touch with friends and family is a major reason they’re there at all.

The second category is the part that was designed, explicitly and with enormous resources, to be hard to put down.

Why this piece of information matters for how you actually use these apps

If the algorithm is the problem, not social media itself, then the goal isn’t to disconnect from the people in your life who happen to be reachable through these apps. The goal is narrower and more achievable: engage with the part that’s communication, and create friction around (or remove entirely) the part that’s an algorithmic firehose.

In practice, this might look like:

This paradigm change is a fundamentally different framing than “spend less time on your phone.” It’s “spend your time on the parts that are working for you, and be deliberate, or absent, from the parts that are working against you.”

That’s a much harder thing to do than it is to say, because right now, both categories live inside the same app, with the same login, and the same infinite-scroll interface design applied to everything. There’s no built-in way to have one without the other.

Which is, frankly, the gap we started Prism to try to close.

If you’ve noticed that certain parts of social media (like messaging friends) feel fine, while other parts (like an endless feed) leave you feeling worse, that’s a normal and common distinction, and it’s worth paying attention to which is which for you.

Sources:

algorithm research intentional-use

Prism helps you open social media with intention, not impulse.