Just one week after Trump “saved” TikTok from a nationwide ban in the United States, two of Silicon Valley’s biggest players — OpenAI and Meta — are making their own grabs for the short-form video-sharing throne, turning entirely to generative AI.
How Are Sora and Vibes Different from TikTok?
While TikTok largely relies on human creators to produce content, Sora and Vibes’ feeds are filled exclusively with AI-generated videos. Instead of users recording themselves, they simply type in prompts and let AI models do the creating. The approach shifts social media from human-driven creativity to, essentially, automated content production, blurring the lines between creator and consumer.
OpenAI’s text-to-video social media app, Sora, was invite-only until October, when it was made publicly available to all iOS users in October. Within 48 hours, it climbed to the top of the App Store, accruing 56,000 downloads on day one, then tripling that the next day. Around the same time, Meta unveiled Vibes, a feature inside its Meta AI app that churns out bite-size digital collages entirely from text prompts. Both companies seem to be testing the same idea: whether artificial intelligence can automate the kind of human creativity that made TikTok a cultural sensation. And so far the results have been mixed, with some users questioning whether these feeds are entertaining or just straight up slop.
The bigger question is whether AI is even capable of producing the same dopamine-driven, binge-worthy experience that has made TikTok so notoriously addictive, or if these apps will wind up being nothing more than a short-lived novelty. The outcome could not only shape the social media landscape, but it could also open up the next frontier for Big Tech’s digital Manifest Destiny, seizing even more territory in our collective attention spans.
What We Know About Sora
Sora is OpenAI’s flagship short-form video platform, built on its upgraded Sora 2 model, which has been programmed to produce more stable motion, lifelike physics and better audio syncing than its original version previewed in early 2024. In a nutshell, the app generates vertical, minute-long clips, merging visuals, motion and sound into lifelike (and often surreal) scenes. Users can depict themselves doing anything from hitting a game-winning homerun in a baseball stadium full of robots to engaging in a Matrix-style, cheeseburger-flinging duel with Ronald McDonald — all they need to do is type out what they want.
When signing up, users are asked to recite a three number sequence and participate in a full face scan, filming a short clip looking at the camera, left, right, up and down. From there, they can create complex scenes, remix videos, add music, adjust styles and even drop themselves or friends into shareable, social-feed-ready cameos featuring their likeness.
So far, a lot of the early feed has been dominated by deepfakes of OpenAI’s very own CEO Sam Altman — serving coffee, frolicking in a field and even getting arrested by police after stealing Nvidia GPUs at Walmart. Users are also testing the boundaries of copyright law, dropping familiar characters into absurd scenarios, whether that’s an arm wrestling Pikachu on the moon or a Nazi-fied Spongebob orating about a “scourge of fish ruining Bikini Bottom.” OpenAI’s current policy allows copyright-protected works to appear unless their rights holders have requested they be removed — a departure from standard fair use procedure that has been raising legal eyebrows.
What We Know About Vibes
Vibes is an AI-powered video creation and discovery platform built within the Meta AI app. It lets users generate short, vertical clips from scratch using text prompts, remix existing ones and layer their own media (photos, videos, music) into surreal, animated scenes.
Unlike Sora, which focuses on hyper-realistic simulations and user cameos, Vibes encourages stylized and narrative-driven content that feels more like a hallucination, emphasizing creative experimentation over photorealism. Take, for example, the clip of a coastal vacation interrupted by roaming dinosaur herds, or the clip of a goldilocked Donald Trump chasing Tom Hanks through a cornfield. The app also then users to cross-post their creations on Meta’s other social media apps, Facebook and Instagram, while surfacing personalized feeds that highlight interaction and inspiration.
Early versions of Vibes will be leaning on Midjourney and Black Forest Labs (two Meta partners) for its underlying technology while the company works on its own proprietary models. Overall, the app is being positioned as part creation studio, part discovery feed in its distinction from competitors TikTok and Sora.
Why Are They Doing This?
The push for generative AI-powered social media apps is likely driven by Big Tech’s race to dominate the next attention economy. If artificial intelligence can endlessly generate content that keeps users engaged, platforms could wind up not having to pay human creators at all. Automated feeds could run around the clock, producing an infinite stream of memes and videos at almost no cost — while still harvesting the same troves of valuable user data that drive ad revenue.
Of course, one could argue that this may eventually erode the “social” nature of social media, leaving us with a seemingly dead internet in which humans consume content but rarely create it. Still, as long as these companies can monetize that attention, there’s little reason to expect them to stop.
It’s also worth noting that OpenAI and Meta aren’t the first companies to try this. Character.ai launched what it calls the “world’s first AI-native social feed” back in 2024, which focused on deep, interactive text adventures rather than video. Google also recently launched a new feature in its Gemini app called Flow, which lets users generate and edit short-form clips made by its text-to-video model Veo. Even TikTok’s parent company ByteDance has its own generative engine, Jimeng AI, which only operates in China right now but could easily be turned into a TikTok spinoff if it added a social component.
Could This Be the End for TikTok?
With more than 1.5 billion active users, TikTok remains the cultural juggernaut that set the standard for addictive short-form video creation in the first place. But its dominance is certainly being tested with apps like Sora and Vibes, which offer the ability to remix or outright AI-generate content that TikTok cannot yet replicate.
While there is no immediate proof that TikTok is on the decline just yet, there’s no denying that generative AI is changing the way we experience the internet for good. In fact, TikTok was ousted from the number one spot in global downloads by OpenAI’s ChatGPT. No matter where the metrics stand today, we’ve already seen this technology shake up everything from entertainment to the way we work, changing not just entire industries but the very rhythm of everyday life. Why wouldn’t it also influence the medium that keeps us glued to our screens for hours on end?
That being said, early public perception of these newcomers has been fairly lukewarm. Some of the top comments on Zuckerberg’s announcement for the Vibes launch read “gang nobody wants this” and “Bro’s posting AI slop on his own app,” with another observing the trend’s potential weak seam: “how is this better than real life.” And Sora has already stirred up some controversy for its sketchy stance on copyrighted material and potential to spread disinformation at a massive scale. For TikTok, this may all turn out to be a competitive advantage in the end.
Only time will tell whether Vibes and Sora are the start of a new era in social media or destined for the same graveyard as the countless other failed apps that have come before them. For now, they will continue to push the limits of what actually counts as online entertainment. AI-generated content can certainly be mesmerizing and engaging, but it’s also hollow — often devoid of any sort of plot, pushing meaningless visuals riddled with odd glitches. If it’s not real, authentic human emotion creativity (or even chaos) that keeps us scrolling for hours on end, these apps may soon reveal what does.
Frequently Asked Questions
What is Sora?
Sora is a text-to-video app created by OpenAI that can generate short, photo-realistic clips based on written prompts. Users simply scan their faces, and they can appear in their own AI-rendered videos with lifelike motion, physics and audio.
What is Vibes?
Vibes is an AI-powered video generation tool created by Meta. Available inside the Meta AI app, it allows users to make stylized, animated short clips from text prompts, or remix existing media, and then cross-post them to Facebook and Instagram.
Could AI-generated content feeds replace traditional social media?
It’s certainly possible, but far from certain. While platforms like Sora and Vibes can generate endless streams of content, they’ve been accused of being nothing more than “slop” machines, incapable of capturing the authenticity and spontaneity that makes traditional social media so engaging. There are also concerns regarding things like copyright infringement and these apps’ capacity for spreading mass disinformation. All of these could offer a competitive advantage for more traditional social media platforms.