Adapted by:
Javier Pastor
Senior WriterComputer scientist turned tech journalist. I've written about almost everything related to technology, but I specialize in hardware, operating systems and cryptocurrencies. I like writing about tech so much that I do it both for Xataka and Incognitosis, my personal blog.
Karen Alfaro
WriterCommunications professional with a decade of experience as a copywriter, proofreader, and editor. As a travel and science journalist, I've collaborated with several print and digital outlets around the world. I'm passionate about culture, music, food, history, and innovative technologies.
Just over a month ago, a disturbing Instagram video went viral. It featured a strange, half-spider, half-man creature in a shopping mall. The video has 3.5 million views and over 23,000 likes, but that’s not the most concerning part.
Overflow of AI-generated videos. The real issue is that this video is part of a flood of AI-generated content overwhelming social media platforms like Instagram and TikTok.
Broken algorithms. According to 404 Media, a unique brute-force attack has disrupted the algorithms of these social media platforms. The attack floods them with AI-generated content, overwhelming their recommendation systems.
AI slop. Brute-force attacks typically involve guessing a password by trying all possible combinations. In this case, spammers are saturating the algorithms to push AI slop, and they’re succeeding. Some now refer to Instagram and TikTok as “AI dumping ground” due to the sheer volume of this content, which overshadows human-created material.
Reality has changed. For many, these platforms serve as sources of entertainment and news. Videos often aim to reflect reality, but that reality has shifted. Many Instagram and TikTok accounts now show little that is real. Some videos are deepfakes that are hard to detect. As users have seen before, the paradox is that they often cannot distinguish between real and fake content.
AI-generated content seeks virality and eventually finds it. Social media creators invest significant time and resources to make their videos go viral, but success is never guaranteed. Spammers creating AI slop, however, don’t face the same challenges. They can produce thousands of videos with minimal effort, flood platforms, and wait for one to catch on. The virality lottery becomes less random when you have vast resources.
Influencers profiting from AI slop. As often happens with such trends, influencers have found ways to profit quickly and easily. One, a 17-year-old named Daniel Bitton, claims to have already earned $2 million. He says, “While others spend 5-6 hours making ONE ‘perfect’ video, we’re cranking out 8-10 shorts in under 30 minutes.” How? By using AI tools.
The “sad hot dog method.” One of Bitton’s friends, a well-known TikTok spammer named Musa Mustafa, uses what he calls the “sad hot dog” method. He explains, “When you’re hungry at 2 a.m., even a sad-looking hot dog tastes BETTER than any Michelin meal that only gives you 2 bites. Well, TikTok works pretty much the same way. Your audience isn’t expecting (or even wanting) perfectly polished videos.”
Mustafa adds, “When’s the last time you saw a viral TikTok and thought: ‘Wow, the color grading on this is incredible!’ Never. Because nobody cares.” In other words, quantity trumps quality.
Social media platforms embrace this content. According to The Guardian, social media platforms aren’t stopping this slop. They’re benefiting from it, accepting it, and even promoting it. They provide tools to facilitate AI-powered content creation, which exacerbates the problem rather than solving it.
One example: Facebook. Meta recently launched a tool for advertisers called Advantage+. It uses AI systems to create multiple versions of an ad, tests them through A/B testing, and selects the best-performing one. For advertisers—and Meta—this is ideal, as it produces more effective ads with less time and money invested.
Are there limits? Some users will undoubtedly reject this type of content. Social media platforms like Bluesky and Mastodon are moving away from algorithm-driven feeds and resemble what Twitter (now X) and Facebook looked like years ago. However, it seems clear that the vast majority of users currently have no issue with AI-generated content. A prime example is AI-generated porn, where impossible combinations—like an AI-generated fish kissing an AI-generated woman or an AI-generated orc marrying an AI-generated girlfriend—are also going viral.
More fuel for the dead Internet theory. For years, people have speculated that the growing presence of bots would eventually marginalize human activity online. Once upon a time, AI-generated texts and images saturated the internet. Now, AI-generated videos and virtual avatars are overwhelming social media platforms. The AI slop is expanding, and the worst part is that neither users—who help propel this content to virality—nor companies—who, as noted, not only allow it but actively promote it—seem particularly troubled by this trend.
Image | shiverchain
Related | The Internet Brought Us ‘Spam.’ AI Is Bringing Us ‘Slop’
Log in to leave a comment