Sam Altman deleted TikTok after discovering he could not stop scrolling, but his decision was not simply about reducing screen time.
The OpenAI CEO intentionally studied the platform while his company was building Sora, its AI video generation platform. His goal was to understand what makes short-form video experiences so engaging and what lessons could be learned from them.
Altman’s experiment was unusual. Instead of simply using TikTok for entertainment, he tried to experience the platform from the perspective of a user.
Speaking on the Relentless podcast, Altman explained that he deliberately used TikTok because he wanted to learn from the platform.
“I made myself get addicted to TikTok, because I wanted to just learn,” Altman said.
What started as a small experiment quickly became more intense. Altman said he began with around 10 minutes of TikTok before bed. Over time, that became an hour, and eventually grew into three-hour scrolling sessions on a Saturday afternoon.
“I was really enjoying it in the moment like a drug,” he said.
After experiencing how powerful the platform’s engagement could be, Altman eventually deleted the app.
Why Sam Altman Studied TikTok While Building Sora
For most users, TikTok is simply a platform for watching short videos. However, behind the experience is a recommendation system designed to understand user preferences and deliver content that keeps people engaged.
Altman approached TikTok differently. He was not only consuming videos; he was studying the experience.
While developing Sora, OpenAI was thinking beyond video generation. The company also had to consider how people would interact with AI-created content and what kind of experiences those interactions could create.
TikTok provided a real-world example of how personalized content can capture attention.
By becoming a user himself, Altman gained firsthand insight into the design choices that make digital platforms highly engaging.
What OpenAI’s Sora Means for AI Video
Sora is OpenAI’s AI video generation platform that can create realistic videos from text prompts. The technology has attracted attention because of its ability to produce detailed scenes, realistic movement, and creative visual content.
However, building AI video technology is not only about creating better models.
It is also about deciding how people will discover, create, and interact with AI-generated media.
Traditional social platforms use recommendation systems to decide what users see next. AI-powered media could potentially take that experience further by creating more personalized content based on user interests.
That creates a challenge for companies developing AI video tools: how do you build an engaging experience without simply recreating the endless-scroll model?
Altman’s TikTok experiment gave him a direct look at that challenge.
The TikTok Lesson Behind OpenAI’s Sora
Altman’s decision to delete TikTok was not just a personal productivity choice. It came from an experiment designed to understand one of the most powerful forces behind modern digital products: user attention.
By experiencing TikTok as both a user and a researcher, Altman saw how quickly engaging content could turn into hours of consumption.
That experience offered insight into the challenges OpenAI faces as it develops Sora and other AI-powered creative tools.
The future of AI video will not only depend on how realistic generated content becomes. It will also depend on how companies design the experiences around that technology.
For Altman, studying TikTok was a way to understand both the possibilities and responsibilities that come with building the next generation of digital media.
As AI-generated video continues to evolve, the biggest question may not only be what AI can create, but how people will choose to engage with what it creates.
