Attention Thieves
Attention and time are today’s currency. Straight Arrow explores how technology is built to capture both, using human psychology to keep us hooked.
Social media is unrecognizable from the era of status updates, “it’s complicated” relationships and “tag yourself” memes. Now, users are more likely to see a carousel of strangers’ divorces than a cousin’s wedding photos, or to read a 400-word personal essay about B2B sales than a status update from a friend “on my way to softball practice!”
Feeds are flooded with Shrimp Jesus, “verified” artificial intelligence “doctors” selling supplements and “viral” trends that might actually serve to gather data for their hosting platforms. An estimated 41% of long-form content on LinkedIn is AI-generated, TikTok and Instagram are rife with undisclosed paid political content and random PR, and platforms continually roll out AI integrations that no one asked for.
Download the Straight Arrow app today to get the stories that matter free from manipulation, bias or agenda.™
Point phone camera here
There’s fake military propaganda, chatbots pretending to be therapists and videos in which AI hosts explain to human users how to make “low-effort” AI videos to “replicate what’s already working in the market right now.” Even group chats could soon have ads.
In 2023, The New York Times reported that the future of social media was “a lot less social.” In 2025, a Reddit user asked, “Where are you supposed to connect with people you know online now?”
The idea that online social networks reflect your real-world social life feels almost quaintly old-fashioned.
Now those networks compete for their users’ attention, serving up AI- and human-generated content designed to attract the maximum number of eyeballs as they try to monetize every click. In short, the platforms exist primarily to turn their users’ time into money.
Marc Smith, a sociologist and the director of the software organization the Social Media Research Foundation, sees platforms as having left their “social” era, when real relationships were the driving force. After a pitstop in what he calls their “parasocial” era — marked by the ascendance of influencers — Smith said he thinks the platforms have entered a “pseudo-parasocial” phase, in which characters on the screen are not even real.
So, he said, it’s time to change the name. It’s no longer social media, he said. “Let’s call it digital media.”
How did this happen?
In the early days of social media, Smith told Straight Arrow, companies like Facebook and Twitter functioned primarily to connect users to their friends and acquaintances through what is known as a “social graph.” There was money to be made in gathering and selling data about how people were connected in the real world.
“Platforms attempt to extract more profit from this service they’re providing,” Smith said, “which used to be access to your friends and your broader social circles, networks, and that stopped working for them.”
Ultimately, connecting with high school classmates and old boyfriends or girlfriends didn’t provide the platforms with enough content to monetize, Smith said. That led to the first major adjustment to their business models.
“People shifted towards the influencer,” he said, “which is sort of the semi-professional content creator.”
Those people, he said, “produce content at a high enough threshold on a particular focus topic that fills what used to be called the news hole.”
But that still wasn’t enough; the platforms needed even more content to keep users hooked.
AI changed everything.
Now, with the growing possibility of an unlimited number of venture-backed AI influencers, digital platforms can keep people engaged with targeted fake content. Smith said this fake content aims to fill the much bigger “scroll hole”: the space in people’s lives that they want to fill with entertainment.
“My sense is, we’re in the infancy of generative AI content,” Carmen Vicente, a social media analyst, told Straight Arrow. “Right now we’re in the flood, right? Like the flood is here. And everyone’s like, ‘Oh, this sucks. We hate slop.’”
But, she said, some people must also like it.
“You get record numbers of people watching Fruit Love Island — why?” she said, referring to an AI-generated shortform series featuring characters like “Watermelina” and “Bananito”.
Vicente sees a difference between the kind of generative AI that tries to confuse or trick people and the kind that can be fun for audiences.
“People don’t like being fooled,” she said. “You don’t like having to look at a piece of content and go, ‘Is that real?’ It’s not a good feeling.”
“I think that what Fruit Love Island did well is that nobody was sitting there watching that saying, ‘Is this real?’” she added. “They weren’t fooled, they were just entertained.”
How are platforms responding to generated content?
Some platforms have tried to remove bot and AI accounts.
As Straight Arrow reported last month, X announced it had culled 42,000 bot accounts. Similarly, Google released a research paper explaining its methodology for culling slop video accounts; it removed 130,000 of them from YouTube that were identified through shared posting patterns.
However, these strategies might be flawed. On July 31, the world’s 270th most-popular YouTube channel, a German science and philosophy account called Kurzgesagt, announced on Reddit that it had been “kicked in the face” by YouTube’s AI removal techniques.
“YouTube’s automatic AI detection tools wrongly think that our very much human-made videos are AI Slop,” a member of the Kurzgesagt team posted. “So it started to choke our channel.”
Meanwhile, audiences began voting with their clicks. Many users moved from social networks to newsletters, and writers are attracted by platforms on which they can directly own their relationships with followers. But those platforms bring their own changes, challenges and tradeoffs — and are not immune to the algorithmic influences that shape both our attention and our information diet, nor the financial influences that change how we relate to the platforms.
Substack recently announced a partnership with the AI company Pangram, giving the latter automatic access to anything posted on Substack, in the name of “AI detection.” While some writers praised the tool, Sam Illingworth, a college professor from the United Kingdom whose Substack has more than 20,000 subscribers, described its introduction as a “witch hunt.”
Substack said it opted to screen for AI content because “we don’t want to wait until your Substack app turns into LinkedIn before we start to learn” how to get rid of junk content.
Following this, LinkedIn added a “Seems Like AI Slop Button.” The response was mixed.
“That feedback helps train the reach algorithm,” a marketing officer said in a blog post. “It is not a content police force. It does not get a post deleted. It does not queue it for human moderation. The poster does not get told. It is only a signal that feeds LinkedIn’s classifiers.”
Smith, of the Social Media Research Foundation, thinks that the profit incentives for platforms to encourage slop remains. To Vincente, the social media analyst, that means that “there will be all sorts of generative AI entertainment that will be super fun to watch.”
Will the internet hold our attention?
It is now no secret that social media companies deliberately designed addictive products. Earlier this year, Meta and YouTube were ordered by a judge in California to pay a woman $6 million in damages after a jury found that the companies’ products had caused her mental harm after she became addicted to them as a child.
Vicente thinks the addictive nature of social media is not going away. Feeds are increasingly tuned to people’s preferences, even while appearing to offer options that allow people to say they’re not interested in certain types of content.
“I mean, if anything, it probably has the opposite effect,” she said. “What they’re not saying is that when you have the power to tune your algorithm, there’s all these other things that come true, which is — one — you’ve just created an even more addictive final product.”
She compared the addictive power of algorithms to the allure of narcotics. Telling a platform what you do and don’t want to see is like saying, “Hey, of all of my drugs of choice, these are the five that I like to get high on the most,” she said. “And so then they go, ‘OK, great, we’ll give you all five of those.’”
So the new wave of generative AI content might make those drugs more potent, she said — and actually make digital platforms more addictive.
Round out your reading
- All your questions about napping, answered.
- Trump’s $5,000 promise echoes past payouts that never materialized.
- Inside the effort to make data centers pay their share of electricity costs.
- Why did the Feds seize the ’largest Martian meteorite on Earth’?
- Photos and video show exactly where and when Trump visited Ground Zero after 9/11.
