COLUMN: Flirting bananas, crying strawberries—welcome to TikTok’s algorithmic gutter

COLUMN: Flirting bananas, crying strawberries—welcome to TikTok’s algorithmic gutter
Credit: Adobe Stock

The algorithm doesn’t show you what you like—it teaches you to like what it shows you

You need to understand one fundamental thing about how TikTok works to grasp why talking fruit can become a global cultural phenomenon. TikTok’s algorithm doesn’t cater to your preferences. It creates them.

The platform’s recommendation system is designed to detect micro-signals of engagement—the length of time your eye lingers on an image, the split-second hesitation before scrolling, the slight slowing of your thumb’s movement. And from these tiny signals, it builds your next feed, brick by brick, like a drug dealer adjusting the dose.

The AI-generated fruit images didn’t go viral because people wanted to see bananas in love. They went viral because the algorithm detected that this type of content generated abnormally high viewing times—and so it flooded the feeds with it.

The Trap of Pleasant Cognitive Dissonance

Why do we find ourselves glued to a crying strawberry? The answer lies in two words: cognitive dissonance. Your brain knows it’s absurd. It knows it’s fake. It knows a banana can’t have a broken heart. And yet, something about the emotional staging—the teary eyes, the trembling voice, the background music—activates the same neural circuits that light up when faced with real human drama.

That’s exactly what makes the phenomenon so insidious. AI-generated content doesn’t ask you to believe it. It simply asks you to feel. And your limbic system—that ancient survival mechanism inherited from a hundred thousand years of evolution—can’t tell the difference between a real emotion and one manufactured by a language model.

And yet, the difference is vast.

Share this post

facebook icon twitter icon linkedin icon
Copied!