The Ghost in the Geranium

If you spend any significant time scrolling through Facebook these days, you’ve likely seen stories about someone like Florence.

Florence is seventy-nine years old. She lives alone on the sixth floor of a building called the Ridgeview Apartments. Her husband, Harold, passed away eight years ago, her daughter lives across the country, and her only daily companion is a stubborn, potted geranium on her balcony that refuses to die.

In a beautifully paced, first-person narrative that recently went viral, Florence tells us about a quiet teenager named Malik who walks home from the local high school every afternoon at 3:15 PM. He walks slower than the other kids, burdened by invisible weights. She waves; he waves back. When Malik goes missing for a week, Florence conquers her arthritic knees, rides an elevator that terrifies her, and climbs three flights of stairs to his dark apartment to find him overwhelmed by his grandmother’s hospital bills and a failing grade in math. Florence doesn’t just comfort him—she coordinates her entire apartment complex. Within days, neighbors are tutoring Malik for free, cooking him chicken soup, and bringing him back from the brink.

It is a masterclass in emotional storytelling. It hits every narrative beat with surgical precision, leaving readers with a lump in their throats and an overwhelming urge to hit the “Share” button.

There’s just one problem: Florence doesn’t exist. Neither does Malik, the Ridgeview Apartments, or the half-dead geranium.

The entire narrative—along with thousands of others flooding the platform daily—is pure “AI Slop.” It is a text string spun out of a large language model, engineered specifically to strip-mine human empathy for cold, hard cash.

Anatomy of the Empathy Bait

This isn’t an isolated creative writing project; it is a highly structured, automated digital business model. If you peek behind the curtain of the pages distributing these stories—Facebook accounts operating under sweeping, sentimental names like Storytellers or Forgotten Chapters of Humanity—a distinct blueprint emerges.

The pages routinely boast tens of thousands of deeply moved, real human followers, yet they follow almost no one in return. They post at a relentless, mechanical frequency, with each account pushing out 10 to 12 long-form vignettes every 24 hours. The accounts generally split their strategies into a few distinct psychological traps:

• The Nostalgia & Authority Hook: Accounts like Forgotten Chapters of Humanity anchor their AI fiction to actual historical figures—Marilyn Monroe, Paul Newman, or Melitta Bentz (the inventor of the coffee filter). They accompany the text with a heavily smoothed, slightly distorted, AI-generated portrait of the figure, weaving a heartwarming story about their past designed to instantly bypass our critical thinking through historical nostalgia.
• The Intimate First-Person Hook: Accounts like Storytellers choose total anonymity to breed intimacy. Introducing a narrator as simply “Miriam, age 74” or “Florence, age 79” creates an immediate sense of authentic, vulnerable confession. The locations are completely generic, the last names are absent, and the details are impossible to fact-check. They speak directly to a very real, modern societal ache: elder loneliness and the desire for generational connection.

The tragic irony of this setup is the comments section. Look beneath these posts and you will find thousands of real, lonely human beings pouring their hearts out. They reply with raw stories of their own isolated apartments, their own deceased spouses, and their own neighborhood dynamics. They are offering genuine human vulnerability to an automated script managed by an entity half a world away.

Follow the Money: How “Slop” Monetizes

When looking at these hyperactive pages, a logical question arises: Why do it? There are no obvious external advertising links breaking up the text. There are no immediate pitches for digital products, dropshipped goods, or miracle creams. The text remains pristine and unbothered by obvious commercial interference. How exactly do these operators convert a fictional story about an old woman on a balcony into a profit margin?

They are playing a calculated, cynical long game. In the digital economy, attention is currency. Tech researchers and investigative journalists have documented the monetization tactics used on social media which drive the explosion of AI slop on platforms run by Meta and Google. Research at the Cyber Policy Center of Stanford University and the Harvard Kennedy School’s Misinformation Review provide examples that show how the process works.

Engagement farmers rely on three distinct monetization mechanics to cash out:

1. Direct Platform Arbitrage

Meta/Facebook actively pays creators for high-volume engagement through its In-Stream Ads and Facebook Performance Bonus Programs. When an elderly user leaves a multi-sentence comment on a post, Meta’s algorithm flags it as “Meaningful Social Interaction” (MSI). The platform rewards this by aggressively pushing the post into thousands of more feeds. Meta then embeds programmatic advertisements inside the comments section or alongside the post on the user’s home feed. The page owner receives a direct financial payout from Meta based entirely on those impressions, without ever needing to direct users off-platform.

2. Audience “Fattening” for Black-Market Page Flips

A Facebook page with 40,000 or 100,000 highly active, real human followers who consistently like, comment, and share is worth thousands of dollars on digital black markets. Spammers use these benign, completely unproblematic, heartwarming stories to quietly “fatten” an audience and build an unpunished account profile. Once the follower count crosses a target threshold, the page is sold on brokers like PlayerUp or Fameswap. Overnight, Storytellers changes its name to a dropshipping retail storefront or a political astroturfing machine, inheriting a massive, pre-warmed audience that completely forgot they ever followed it.

3. The Delayed Link-Farm Edit

Facebook’s algorithm severely penalizes posts that contain outbound links right out of the gate, as the platform wants to keep users trapped inside its ecosystem. To circumvent this, spammers execute a classic bait-and-switch. They wait for a wholesome post about an elderly person or a historical figure to gain viral momentum. Once the post racks up 50,000 shares and establishes a massive algorithmic footprint, the operator quietly edits the main caption or drops a pinned comment reading: “If Florence’s story touched your heart, click here to read our full collection of uplifting tales.” That link redirects users to an off-platform clickbait blog blanketed in high-paying programmatic ad blocks.

Behind the Digital ID: A deep dive into the public “Details” section of these exact pages exposes the identity layer. The Storytellers and Forgotten Chapters of Humanity pages list the same person as their contact—“Hemantha,” which is a distinctly Sri Lankan name. “Hemantha” also owns numerous other Facebook pages with names such as Daily Story, American Philosophy and Unknown Tales that Shaped Humanity. The different Facebook accounts give Hemantha’s address variously as Texas, North Carolina and California, and in the case of Storytellers, the address is a jumbled mix of North Carolina and California. This is a classic hallmark of overseas click-farms using compromised or synthetic US identities to bypass Meta’s geographic monetization restrictions—allowing overseas operators to tap into the much higher payout rates that Facebook reserves for US-based traffic.

The Dilution of the Digital Commons

What makes this trend profoundly unsettling isn’t necessarily the financial deception; it’s the systematic automation of empathy.

The text strings generated by these models are explicitly engineered to weaponize human kindness, community spirit, and tears. They take our absolute best impulses—our desire to see the overlooked, to protect the vulnerable, and to rally around our neighbors—and turn them into data points for an algorithm.

When our digital spaces become heavily saturated with hyper-optimized, artificial human connection, we risk developing deep, defensive psychological scar tissue. We begin to view genuine stories of local triumph, authentic community organizing, and real human suffering through a lens of permanent suspicion.

The true danger of the “empathy farm” is that the next time a real, flesh-and-blood Florence calls out from a balcony in our local community, a public that has been burned by a thousand synthetic geraniums might just look down at their phones, assume it’s slop, and keep right on walking.

The Takeaway

I’ve been reading about and occasionally writing about propaganda for a long time. The bad news is that people are not very good at telling the difference between real versus fake information, and the techniques used to manipulate our attention and opinions keep developing and becoming more sophisticated. The good news, however, is that people are also working to become more discerning. Here are a few resources that you might find useful:

• How to Combat Misinformation, by FactCheck.org
• Disinformation Stops With You, by the Cybersecurity and Infrastructure Security Agency
• How to Spot Fake AI-Generated Facebook Stories

Guard your attention online. If a story feels perfectly engineered to tug at your heartstrings but contains no verifiable entities, specific locations, or real-world accountability, don’t feed the machine. Don’t comment. Don’t share. Save your waves for the real people walking slower than the rest on your actual street.