Newsweek/Getty
By Anna Rahmanan
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
Perez Hilton was alone, bloodied and visibly distressed on TikTok Live as he appeared to harm himself in front of an audience that could neither look away nor reach through the screen. It was horrifying to watch: a human emergency unfolding in the same vertical frame normally reserved for gossip, jokes and disposable entertainment.
After viewers made multiple emergency calls, Miami-Dade deputies went to Hilton's home, found him alone and worked to de-escalate the crisis before he was transported to a local hospital Tuesday night. Authorities have not publicly disclosed the longtime celebrity blogger's current condition.
The footage was terrifying for multiple reasons. Hilton's apparent distress was shocking enough, but the episode also exposed how profoundly social media has changed—and how much our conversations about it have failed to keep up. For years, we debated how platforms were reshaping society: shortening attention spans, fueling comparison, spreading misinformation and rewarding outrage.
Read More on News
Those conversations belonged to the first era of social media, but we have since entered another. Today's platforms don't simply influence how we think or communicate; they erase the distance between private crisis and public spectacle, turning viewers into consumers, first responders and bystanders all at once. Those watching Hilton weren't just witnesses—they had to decide whether to call for help, keep watching or scroll away as a human emergency unfolded in the same feed as comedy sketches, beauty tutorials and advertisements.
The internet once required users to seek out specific disturbing images and videos, but algorithms now deliver them automatically, packaging someone's worst moment in the same format as every other piece of content. What can look like a loss of empathy may actually be something more unsettling—platforms have rewired our relationship to human suffering itself, making even a real-time mental health crisis feel like just another thing to watch before the next video begins.
The Old Internet Made Us Search for Spectacle
Let’s be clear: Millennials hardly grew up in an innocent digital world. Message boards, file-sharing sites and unmoderated corners of the web gave children access to violence, pornography and humiliation long before adults understood what broadband had brought into their homes.
Hilton’s celebrity-gossip site belonged to that unruly internet. He became famous in the 2000s by turning celebrities’ private failures and public embarrassments into participatory entertainment. Readers of his posts helped to establish another person’s humiliation as a communal online event.
Related Story
Trump Sells Truth—And Buys Himself a Leash
But much of the social internet millennials used was still organized around deliberate choices. MySpace users, for example, built their own networks and Facebook began with friends and classmates. MSN Messenger required a contact list and early Instagram largely showed posts from accounts a user had chosen to follow.
Those spaces could be vicious, but they were comparatively easy to leave. They did not continuously study a user’s behavior to determine which emotional stimulus might keep that person watching.
TikTok—and the current versions of Instagram and Facebook—reversed that relationship.
Recommendation feeds infer interests from signals that include viewing behavior, so lingering over something disturbing can influence what appears next. A 2026 study of TikTok’s recommendation system found that users could shape their feeds but sometimes struggled to prevent the platform from returning to unwanted subjects. Researchers also found that TikTok’s strongest rejection tool, “Not Interested,” was buried in the interface.
Jessica Hamilton, associate professor of psychology at Rutgers University, noted that engaging with self-harm content can also shape what users see next.
"Engaging in self-harm content on social media (and reaction videos to it) will train the algorithm to increase exposure to these videos," she said to Newsweek. "People should avoid watching these videos (viewing/commenting/sharing, etc), especially if you feel distressed by this content or think you may engage in self-harm."
No one has to go looking for a crisis anymore. The crisis arrives on its own, slipped between a comedy clip and a skincare tutorial.
The Feed Turned Viewers Into Witnesses and Responders
Hilton’s livestream forced viewers to confront a question that earlier social platforms rarely posed so directly. What does a person owe someone in crisis on the other side of a screen?
Some viewers called emergency services and those calls helped authorities find Hilton. Their response complicates any easy claim that online audiences have become indifferent because, for them, watching did not necessarily mean enjoying, endorsing or ignoring what was happening. In this case, people in the audience became an emergency alert system.
It is also possible that Hilton’s decision to broadcast an intensely private crisis helped bring assistance to him. A livestream can place witnesses inside an emergency that might otherwise unfold unseen. Social media has connected vulnerable people with support, information and communities, and the American Psychological Association has cautioned that its effects vary according to the user, the content and the design of the platform.
The same mechanism creates the central paradox of Hilton’s broadcast—the people helping him were also part of the audience watching him collapse. Others remained without intervening. Clips were then copied, detached from the emergency itself and circulated to people who could do nothing except consume them.
The interface made rescue and spectatorship look almost identical as both required viewers to remain on the stream and both registered as attention.
Younger Audiences Inherited a System That Rewards Watching
Younger users are the first generation to grow up inside platforms where the boundaries between witnessing, helping and consuming can collapse during an ordinary scroll. It's tempting to describe them as desensitized, but that misses the larger transformation. In fact, the defining feature of today's social media isn't that it numbs people to suffering, it's that it packages suffering like everything else.
According to the Pew Research Center, roughly one in five American teenagers reported using TikTok almost constantly in 2025, describing a generation for whom algorithmically curated feeds are less a destination than the backdrop of everyday life. What appears in those feeds may matter as much as the time spent on them, basically. One small longitudinal study found that adolescents who encountered self-harm content during an eight-week period were more likely to report urges and behaviors related to nonsuicidal self-injury, though the researchers did not find the same association with suicidal thoughts and cautioned against broad causal conclusions.
“In some ways, seeing self-harm happen in a public space and more often, such as on social media, can desensitize them to the violence,” Hamilton said. She added that repeated exposure can also reduce empathetic responses, reinforce stigma and make it harder for young people who self-harm to seek support.
Repeated exposure makes moments of profound distress feel structurally indistinguishable from everything around them. A mental health crisis arrives in the same frame as a makeup tutorial or recipe, demanding the same swipe, pause or tap. Younger users didn't build that system, but they inherited platforms designed to optimize attention above all else.
Mitch Prinstein, a psychology professor at the University of North Carolina at Chapel Hill, told Newsweek that “without regulations on social media, kids are seeing a whole lot that we used to protect them from” on television and in movies. He added that “recommendations, likes and algorithms are making us react differently to what we see, even at the neural level.”
The older internet taught users that horrible things existed online and the current one measures how long they look.
Perez Hilton Became the Spectacle He Once Published
Hilton’s place in this history becomes almost ironic when measured against the changing landscape.
He built a career during an era when celebrity breakdowns fueled the internet, as the lowest moments of figures like Britney Spears, Lindsay Lohan and Amy Winehouse became around-the-clock entertainment for blogs, magazines and television. Hilton has since expressed regret for the role he played in that culture, but remorse cannot erase the machinery it helped build. Tuesday night simply flipped the script, placing him on the other side of the screen and revealing how much that machinery has evolved beyond him.
"There has long been a commercial market for vulnerability—think reality TV, tabloid talk shows and (perhaps ironically) celebrity gossip culture," Cornell University professor of communication Brooke Erin Duffy said to Newsweek. "Given attention economy logics and platforms' push to extremes, social media culture seems to operate as a vulnerability economy."
The difference is that platforms don't merely amplify public humiliation, they algorithmically distribute it.
TikTok says it prohibits content that promotes or normalizes self-harm and relies on both automated systems and human moderators to enforce those rules. Yet Hilton's livestream remained online long enough for viewers to watch it, report it, record it and redistribute it. A system designed to maximize attention will always face the same question: can it reliably distinguish between compelling content and a person in immediate crisis?
Research may eventually tell us more about how recommendation algorithms shape behavior and which design changes reduce harm. But Hilton's livestream has already exposed the moral dilemma of this moment: the internet once taught us to look at other people's pain, today's platforms train us to keep watching it, leaving viewers to decide when watching becomes helping, when it becomes harm and whether they can still tell the difference.
View 0 Comments
Submit Correction
View Editorial & AI Guidelines