Brain Rot Pays: How Social Media Cashes In on Your Lost Time

How traditional social media continues to rot your brain—and why its new safety features deserve a closer look.

You open Instagram to answer a message. Twenty minutes later, you know that a stranger has bought a Lamborghini, another stranger hates your generation, and three people have discovered the secret to happiness.

You still haven’t answered the message.

Somewhere, this visit counts as a success. You stayed. You watched. The system found more things you would have trouble ignoring.

Whether you got what you came for is a different question.

That gap is the engagement game. In 2026, we have better research about its costs, more internal company documents to examine, and some improvements worth taking seriously. We also have plenty of reasons to check the improvements before applauding them.

A business can succeed while its customers lose

Imagine a restaurant judged entirely by how long people stayed at the table. A wonderful dinner would count as a success. So would a customer who couldn’t find the exit.

The stopwatch wouldn’t know the difference.

In January 2026, Meta reported that changes to its recommendations had produced a 20% increase in time spent on Threads during the previous quarter. It also celebrated more Facebook views and extra time on Instagram from AI-translated videos. These are Meta’s own figures for improvements made in late 2025. Meta’s performance update

Perhaps people found better conversations. Perhaps they found a more efficient way to waste an evening. Time spent, by itself, cannot settle that.

This is the trouble with treating attention as proof of satisfaction. People can keep doing something they wish they could stop doing. Anyone who has read a comment fight to the bitter end knows the feeling.

AI adds more material to the menu. Meta’s Vibes feed, introduced in 2025, lets people create and remix short AI-generated videos. There are creative possibilities here. There is also another supply of things to watch. Vibes announcement

The feed can keep finding something else to show you. Your evening still has an end.

Sleep is part of the market

Once a company has persuaded you to open its app, growth can come from persuading you to stay longer. Those extra minutes have to come from somewhere. Another app, a conversation, the book beside your bed. Sometimes, sleep.

The bluntest version of this argument predates the current social media debate. In July 2017, an analyst asked Netflix’s then-CEO Reed Hastings how the company would maintain and expand its leadership. His answer was: “More watching, less sleep.” Netflix is a streaming company, and this is an old public statement, not a new social media leak. It nevertheless puts the competition for our time into unusually plain language. Netflix earnings transcript, page 14

For a social media example, California’s 2024 complaint against TikTok quotes an executive acknowledging that the app could crowd out sleep, eating, movement, and face-to-face contact. The executive was warning about the consequences. That shows awareness of the trade-off; it does not establish an instruction to make users sleep less. California’s TikTok complaint, paragraph 2

Even without a public declaration, the commercial incentive is hard to miss. For an advertising-funded platform chasing more viewing time, keeping someone awake for another video offers another chance to show an ad. My bet is that, without a strong reason to protect bedtime, competition for attention will keep pushing into it. The company can collect the benefit of those extra minutes while the user carries the cost of losing sleep.

That is an argument about incentives, rather than proof of a secret plan. User trust, regulation, and the long-term value of a satisfied customer can pull the other way. But protecting sleep requires accepting that some viewing opportunities should stay lost—and judging a product team by something more useful than how many of them it captures.

What we know about that scrambled feeling

“Brain rot” is the everyday description: you put the phone down feeling scattered, restless, and strangely tired after doing very little. Claims of permanent brain damage go much further than the research supports.

The evidence is more interesting when we stop demanding one explanation for everything.

Feeds, Feelings, and Focus, a large 2025 review, pulled together 71 studies involving roughly 98,000 people. More short-video use went together with worse attention and a harder time resisting impulses. It did not prove that watching videos caused those problems. People already struggling to concentrate might also be drawn to a feed that asks very little of them.

The study also found a link with poorer sleep quality, while the link with self-esteem was unclear. These numbers describe associations, not percentage losses or proof that scrolling caused the outcomes.

Taming the Endless Scroll?, published in 2026, found a similar worrying pattern among younger people. But most of the studies were snapshots of their lives, which makes it difficult to work out what caused what.

Another review, Short-form Video, Cognitive and Mental Health Outcomes, added something useful: how people use the feed may matter more than a simple count of minutes. Watching something enjoyable can improve your mood. Searching for help, talking to friends, and scrolling because you cannot bear another quiet moment are different experiences.

And Social Media Exposure and Anxiety and Depression in Young People found only small links between general social media use and depression or anxiety in young people. Its authors argued that social media use was unlikely to explain the youth mental health crisis. That finding belongs in the story too.

The claim that every minute online is poisoning a generation makes a convenient headline. It gives us very little help in deciding which parts of the product to change.

The more useful place to look is where people lose control: the visit they meant to end, the content they keep seeing against their wishes, the bedtime that keeps moving.

The conversation inside the company

Publicly disclosed internal documents bring the argument closer to the people making the products.

In a Meta employee conversation from January 2024, filed in court in March 2026, one participant described ranking a child’s Facebook feed by what would make them consume the most content. Employees also discussed reviewing teams’ targets and finding alternatives they could defend in public. Their names are redacted; the published record is a transcript of an internal discussion, rather than a company policy statement. The conversation and court reference

The revealing part is the disagreement itself. Some people inside the building could see the problem with using consumption as the goal.

Another internal report, dated April 2024 and filed in May 2026, recorded teen adoption of Instagram’s Daily Limit tool at 1.4% and Take a Break at 4.4%. Those are old figures, before newer protections. They nevertheless illustrate how little the existence of a safety button tells us about its reach. The time-management report

A company can point to a control. A parent can assume the problem has been addressed. Very few users may have switched it on.

Then there is Project Mercury. In a legal brief made public in November 2025, school-district plaintiffs alleged that Meta stopped research after people who temporarily deactivated Facebook reported feeling less depressed, anxious, lonely, and prone to comparing themselves with others. The plaintiffs’ brief, page 27

Meta said the research was stopped because its methods were flawed. The full study was unavailable for independent assessment in the sources reviewed here, so treating the allegation as a settled scientific result would be dishonest. Meta’s response, reported by Reuters

Publish the methods and the results. Let people outside the company examine them. A disagreement about research should be open to research.

A healthier feed, or better packaging?

There have been improvements. Dismissing all of them would be as lazy as believing every announcement.

I have a small example from my own experience with YouTube. I open the three-dot menu on the Shorts section and choose “Show fewer Shorts.” For a short while, it seems to help. Then the Shorts start crowding the feed again. The effect of asking for less feels remarkably short-lived.

That is my experience of my own feed, not a measured expiry time for everyone. YouTube’s instructions for managing recommendations document the control, but do not say how long its effect should last. They describe fewer recommendations, rather than a permanent removal of Shorts.

Still, it leaves me with a reasonable question: how often should I have to tell the same product that I want less of something? A preference that needs constant renewal starts to feel like a negotiation I never agreed to enter.

There is a more serious example in the TikTok record. NPR’s reporting on Kentucky’s 2024 lawsuit described internal material about time-management reminders being assessed partly through their effect on public trust. The state alleged that the tools were serving the company’s image rather than meaningfully reducing use. TikTok defended its safeguards and pointed to parental controls and privacy protections. These are disputed historical allegations, but they give us a concrete reason to ask what a safety feature is being measured against. NPR on the internal TikTok documents

In June, Meta announced wider availability of more restrictive default content settings for Teen Accounts. It was also testing ways to stop teenagers seeing the same kinds of sensitive content repeatedly. The company reported encouraging results from an outside safety assessment it had commissioned. Teen Accounts update

Keeping unwanted material out of a teenager’s feed is a worthwhile result. It still leaves the question of whether that teenager can put the phone down. A test of content filtering cannot answer a question about sleep or compulsive use.

Threads has also introduced Dear Algo: a way to ask publicly for more or less of a topic, with the adjustment lasting three days. It launched in four countries in February. Dear Algo announcement

There is value in telling the system what you want. But choosing the subject gives you limited control over the game. A feed full of your favorite things could be harder to leave. That is a possibility to test, rather than an outcome this feature has already demonstrated.

The August agreement between Meta and US attorneys general goes further. For teenagers in participating jurisdictions, Meta committed to a combined default two-hour daily limit across Instagram and Facebook that requires parental permission to disable, overnight restrictions, and options to turn off autoplay and make a feed without personalized recommendations the default. Meta’s account of the agreement

These changes deserve more attention than another wellness tab. They can put limits directly into the experience instead of waiting for someone to find a setting.

Still, the announcement is the start of the test. It leaves practical questions about identifying teenagers correctly, bypassing restrictions, and moving the same habit to another app. Protections limited to teenagers also leave adults playing much the same game.

I would judge every improvement by three fairly ordinary questions:

• Who gets it? A useful setting buried in a menu may protect almost nobody.
• What changes? Look for fewer visits people regret, less unwanted content, and less scrolling after bedtime.
• Who checks? Publish enough information for people outside the company to examine whether it works over time.

A real improvement can be small. A cosmetic improvement can have an enormous launch campaign.

The difference will show up in people’s lives, if somebody bothers to measure it.

The feed is finally part of the argument

Regulators are looking at this too. In February, the European Commission issued preliminary findings against TikTok over features including infinite scroll, autoplay, notifications, and its recommendation system. It questioned the value of time-management tools that were easy to dismiss. February findings

In November, it issued further preliminary findings about recommendations repeatedly drawing people toward themes such as sadness, isolation, and body image. These are preliminary findings, with the process still to run. November findings

That brings an overdue question into view: what happens when the system keeps serving something a person is having trouble turning away from?

A platform needs a better answer than pointing to a button somewhere in the settings.

Who lives with the consequences?

A rise in time spent can appear in a company’s performance report. The user lives with whatever filled those minutes, and whatever got pushed aside.

Those outcomes can be good. A friendship, a useful idea, a customer for a small business. They can also be an unfinished task, another lost evening, or a habit the user wishes they could break. The interests overlap often enough to keep us coming back. They can also pull apart.

That is why I want to know whether a platform would keep a change that helped people leave satisfied, even if it meant fewer opportunities to show them ads.

Web3 social: build for the person who closes the app

If you are building a Farcaster client, a Web3 community, or a new social app, this is your opportunity to change the incentives. Decentralization gives us design choices. A token can still give someone another reason to keep checking the screen.

In my earlier article on social media’s engagement traps, I argued for intentional use, user control, and meaningful contributions. In Unlocking Human Potential Through Social Media, I explored learning, mentorship, and connections beyond the screen. Those ideas give us five practical starting points:

1. Give the feed an ending. Build natural stopping points. Let someone catch up, find what they need, and leave. Treat a completed visit as a product success.
2. Make user control real. Let people control their data and understand how their feed is selected. Honor requests to see less, offer different perspectives, and make privacy choices clear. A preference should survive the next visit.
3. Help people discover what they can do. Connect an interest to a course, a mentor, or a small project. Count the skill someone develops and the useful help they receive among the outcomes worth pursuing.
4. Turn online connections into life outside the app. Help communities arrange workshops, local meetups, and accountability partnerships. Make it easy for a good conversation to become something people do together.
5. Reward contributions worth making. Recognize useful answers, creative work, and help given to others. Be careful with rewards for posting volume, check-ins, and constant activity: a token can pay people to repeat the same bad habit.

Choose one of these commitments, build it into the default experience, and test it with your community. Ask people whether they accomplished what they came to do. Publish what improved and what failed. If you are a creator or community organizer, ask the tools you use to meet the same standard.

The next Web3 social success story could be a person who finds a mentor, makes a friend, learns something useful, and closes the app satisfied. It's on us to build products that can count that as a win.