AI Will Not Eat Social Media
Everyone is wrong about AI Slop #2
Generative AI is poised to reshape social media dramatically—no question about it (is anyone else super mindful of using “—” these days?). From content creation to recommendation algorithms, its influence will be far-reaching. But contrary to the more alarmist takes, I believe that AI will not “destroy” or “eat” social media as we know it. Especially the kind of AI-generated content that exists today. This prediction overlooks what we know about people’s social media experiences and the creator economy. Namely, they overlook three things, which I discuss in the essay.
Authenticity is the currency of social media creators
The idea that AI will overtake social media assumes that users will be satisfied consuming content generated entirely by machines. This is highly unlikely, considering how influencer-driven social media platforms like TikTok and YouTube are currently. Influencers thrive because users develop parasocial relationships with their creators, they follow their lives, care about their opinions, and participate in their narratives. The keyword here is “authenticity,” which AI slop is anything but.
But how about AI influencers? Yes, one could imagine a world where people form connections with fictitious characters, similar to how Japan’s virtual idols like Hatsune Miku have gained devoted fanbases. But this kind of content is carefully crafted. Miku isn’t just a product of automation; she’s the result of intentional worldbuilding, aesthetic curation, and human creativity.
The “AI will eat social media” discourse implies that sheer scale will tip the balance. We can now generate endless amusing content! But scale is at odds with authenticity. Even in the absence of the very real anti-AI sentiment that exists online, the fact that content is produced en masse, without exceeding care or thought, will make it very hard for people to create the emotional connections that drive the online spaces.
Social media users are not passive consumers of content
Social media is not a one-size-fits-all broadcast model. It’s a sprawling, fragmented ecosystem made up of micro-niches, subcultures, fandoms, and highly individualized streams of content. Unlike traditional media, which pushes the same content to everyone, social media is shaped by the cumulative actions and preferences of each user. People don’t just consume content. They curate their own feeds through what they click on, follow, like, ignore, comment on, and share. In doing so, they effectively place themselves into specific corners of the internet, each with its own norms, aesthetics, values, and inside jokes.
These corners aren’t just algorithmically generated—they’re socially constructed. BookTok isn’t just “videos about books,” and Skincare YouTube isn’t just “product reviews.” These niches become communities, defined by shared language, common reference points, and a sense of belonging. Participation requires a kind of cultural literacy. You have to know how to speak the dialect, what trends are current, what behaviors are rewarded, and what signals authenticity.
Users build their content diets through a potpourri of “niches,” and it is just implausible that GenAI will eat all of them. In some “types” of content, I’d argue that using GenAI may be more of a liability than a disadvantage. Sure, AI will likely be used as a tool to create memes, but will it be helpful for content rooted in lived experiences?
Social media is built to handle enormous amounts of content
The “AI will eat social media” narrative treats scale as something novel or destabilizing, when in reality, social media platforms have been built to handle overwhelming volumes of content. TikTok, YouTube, Instagram, X, and Reddit are not overwhelmed by too much content; they are defined by it. These systems are designed from the ground up to process, sort, and surface content in real time. Most of what gets uploaded—whether human-generated or AI-generated—is of low quality, ignored, and quickly buried. That’s business as usual.
The fact that people can now use AI to generate large volumes of content doesn’t fundamentally change how these platforms operate. If users don’t engage with it, it simply won’t circulate. And if they do engage with it, then sure, platforms will show them more of it. But they won’t show only that kind of content, because as discussed earlier, users aren’t passive consumers. They seek out relevance, novelty, personality, and cultural context. A wall of AI-generated “dank memes” might go viral, but it won’t define everyone’s social media experience.
GenAI will almost certainly reshape the texture of social media, but it won’t erase the dynamics that make these platforms compelling to begin with. Social media is shaped by human behavior, taste, identity, and community. Authenticity still matters. User agency still matters. And the infrastructure is already built to filter the flood. AI might change the how of content creation, but it won’t change the why of engagement. For better or worse, social media is still powered by people—and that’s not going away anytime soon.