Author| Xing Fei
Editor| Liu Jingfeng
Game-changers are always the ones who venture into uncharted territory and carve out new worlds.
When OpenAI unveiled the demo for its next-generation AI video model, Sora2, the tech world was stunned.
It can generate a 10-second clip almost instantly — complete with convincing physical realism and cinematic multi-camera shots — bringing it astonishingly close to the level of professional human creators.
Soon after, Dongguan Securities released a report noting that “the launch of Sora2 and its companion social applications marks the convergence of AI video generation and social interaction. This could reshape how content is created and distributed, potentially signaling the ‘ChatGPT moment’ for AI video.”
That analogy hit the industry’s nerve. ChatGPT represented the turning point when incremental progress in AI became a quantum leap — proof that massive models, vast datasets, and immense computing power could unlock something fundamentally new.
What’s even more striking is that Sora2 isn’t just a video generator anymore — it’s a complete revolution. Entire ecosystems connected to video — film and TV, social media, e-commerce — are on the verge of being rebuilt, with many business models up for reinvention.
And for everyday people, one question looms largest: how will Sora2 change the way we make money?
01 Sora2 created a new Global Business Model
Only 4 days after its launch, the Sora app shot to the top of the free apps chart on Apple’s U.S. App Store — surpassing both Gemini and ChatGPT — making it a true “instant sensation.”
According to estimates from Appfigures, the iOS version of Sora reached a total of 164,000 downloads within just two days of release, with 56,000 downloads on launch day alone. (Image caption: App Store screenshot)
Sora’s meteoric rise can be attributed to two groundbreaking AIGC-driven social features it introduced: Cameo and Remix.
Cameo is built on OpenAI’s ongoing work in world simulation models. With just a short video clip, users can “transport” themselves or their friends into virtually any setting — from the Colosseum in ancient Rome to a futuristic cyberpunk city or even a video game world — where they can move and interact naturally within the generated environment.
Remix, on the other hand, is a core interactive mechanic in Sora. When users see a video they like, they can tap the Remix button to create a new version based on it. For example, after watching a clip featuring Iron Man and Spider-Man, a user might enter a prompt like “add me into the scene,” and Sora will generate an entirely new, related video — now showing the user interacting with Spider-Man. This dramatically lowers the barrier for joining viral trends and storylines, making AIGC participation effortless for everyone.
Another reason for Sora’s explosion in popularity is OpenAI CEO Sam Altman himself, who personally joined in by releasing his own likeness rights — accelerating the app’s viral spread. On international social media, Altman has been spotted (virtually) stealing GPUs from a store, dancing in front of the pyramids, sparring with Bruce Lee, and playing basketball with Kobe Bryant — all rendered so convincingly that they could easily pass as real footage.
Meanwhile, on Red Note（Xiaohongshu）, quick-moving creators have already featured Sam Altman in livestream sales, where he fluently delivers lines like “Three, two, one — link it now!” — and somehow, it feels perfectly natural.
Picture from Red Note
Thanks to the seamless quality of its generated videos, Sora2 is opening up new possibilities across a wide range of industries.
Take e-commerce as an example. OpenAI is building a closed-loop business ecosystem that connects its tools and partners: Sora stimulates consumer demand through content, Stripe provides secure payment infrastructure, and Shopify/Etsy supply the products. As video-driven shopping and livestream e-commerce gain global traction, Sora—through the Agentic Commerce Protocol (ACP)—could usher in a new era of end-to-end AI-driven commerce.
OpenAI also appears to be doubling down on its e-commerce capabilities. Just days ago, the company announced a new Instant Checkout feature, enabling users to complete purchases directly within ChatGPT conversations. This marks a deeper integration of OpenAI into the e-commerce ecosystem, transforming ChatGPT from a product recommendation tool into a complete shopping and payment channel.
In the future, if Sora-generated product showcase videos, buyer review compilations, and other creative assets are linked into this same commerce loop—and integrated with APIs from major e-commerce platforms—OpenAI could establish a new “people–goods–context” model, distinct from traditional e-commerce logic.
At the same time, marketing and production costs could drop dramatically. In the past, high production expenses, long turnaround times, and professional barriers made large-scale expansion difficult for sellers. Now, those challenges are rapidly disappearing.
Peng Shengcai, CEO of Agent Platform firm Shuxin Technology, notes that tools like Sora2 can make previously impossible business models viable—or even create entirely new ones. For example, launching a product overseas used to involve finding agents, managing distribution channels, and building brand awareness step by step. Now, with AI and related technologies, companies can streamline the entire process.
We’re already seeing a new generation of startups leveraging China’s strong supply chain and manufacturing base, combined with AI-powered tools for marketing, ad placement, and customer service, to bring products directly to international markets. This represents a radical departure from traditional export models—some “one-person companies” can now manage the entire operation. For these businesses, AI acts like a set of wings, making once-impossible ventures achievable and previously unreachable markets accessible.
However, this AI-led commerce model also raises the bar for brands. According to Peng Shengcai, the old paradigm was “channel is king”—brands competed to cover as many sales channels as possible, and platforms played a crucial role in visibility. But with AI, consumers may simply describe what they want in natural language, and the AI handles the rest—matching them directly to the right product or service. This shift reduces the importance of channels and amplifies the importance of brand value.
“In this new model,” Peng explains, “the only way to secure your place in the AI-driven consumer ecosystem is to build a strong brand. In the past, we spent huge sums on marketing and ads. Now, all we need is to make great products and services, distribute them through the MCP (Model Context Protocol), and the AI will automatically find the right customers. It’s a transformative change—but in the long run, it will lead to a healthier, more sustainable industry ecosystem.”
(Note: MCP, or Model Context Protocol, is an open standard for connecting AI applications with external systems. Through MCP, AI tools can access data sources, workflows, and utilities—allowing them to retrieve key information and execute complex tasks.)
02 Flipped the short video industry on its head
Even more disruptive is how the Sora app, powered by Sora2, adopts a TikTok-style swipe interface instantly lowering the barrier for consumer use.
Observers widely agree that Sora has effectively “flipped the short video industry on its head.” Its impact could pose a serious challenge to the business models of advertising, celebrity endorsements, film and television, and the creative industry at large.
In September from this year, TVB released a short drama titled In <My Heart, You Are Unique> — a production 100% generated by AI. Every element from the lead actors and background extras to the sets, soundtrack, and even the editing — was created by AI. In other words, throughout the production process, AI took on every major role: screenwriter, director, art designer, editor, and post-production team. It even played a key role in marketing and content distribution.
From a technical standpoint, Sora can now generate short videos up to 25 seconds long (15 seconds for standard users on mobile/web, 25 seconds for Pro users via web), with near-4K visual fidelity. The output is sharper, smoother, and features improved frame-to-frame continuity, along with multi-language synchronization and translation. Google’s Veo2, after its recent upgrade, can also deliver 4K cinematic-quality videos of over two minutes, showing how these tools are driving a quantum leap in content production efficiency and quality.
As AI becomes deeply embedded in every stage of short drama production, it is rapidly becoming the core engine for cost reduction and productivity gains. On the production side, AI now handles everything from scriptwriting and storyboarding to virtual filming and smart editing, drastically shortening the time from concept to completion. On the operations side, AI can generate massive amounts of marketing materials for A/B testing, achieve precise ad targeting, and leverage user data analysis to refine content continuously — creating a self-optimizing creative cycle.
Traditionally, a short drama required 2–3 months of production time and a trial-and-error budget of ¥500,000–2 million RMB — already an industry-optimized baseline. With AI-driven workflows, however, production time has dropped to 10–15 days, and costs have fallen to ¥60,000–150,000 RMB, cutting up to 90% of the expense per episode.
This dramatic shift is reshaping the fast-growing short drama industry, unlocking new creative and commercial possibilities.
At the same time, Sora’s influence extends far beyond media — into social interaction itself. Its swipe-to-browse interface has earned it the nickname “the AI version of TikTok.” Users can upload photos and prompts to generate videos, or use the Cameo feature to insert themselves or friends into AI-created scenes. This highly interactive, personal experience gives Sora a strong social DNA.
Once the “browse → inspire → create → share” loop begins, users become deeply engaged in the ecosystem. Every new video fuels the platform’s growth — expanding its creative database and evolving into a self-sustaining “content flywheel.”
In this sense, “AI + social” could well become the dominant format of the next digital era.
Sora2’s impact on the advertising industry has also drawn widespread attention — though so far, its disruptive power hasn’t fully materialized.
Peng Shengcai offers an example: in advertising and marketing, the creation of copy, images, and videos happens at the very end of the production chain. That end stage typically involves an entire team — with planners, designers, directors, and editors all playing different roles. Sora2 might affect the video creators or editors, but within an advertising supply chain that has evolved over more than a century, this segment represents less than 1% of the overall process. That’s why, as Peng puts it, “Sora2 alone isn’t likely to disrupt or reshape the entire industry chain.”
That said, Sora2 will certainly lower costs and boost efficiency, allowing some people to reap short-term rewards. For example, even if advertisers keep paying the same price for video ads, those who master tools like Sora2 can produce content 10 to 100 times faster. With lower production costs but the same selling price, profit margins naturally expand.
However, Peng warns that this advantage will be temporary: “As more creators adopt AI, supply will skyrocket, and competition will inevitably intensify. If someone can deliver 10 videos at today’s price, others will soon offer 50 or even 100. So yes, there’s a brief window of opportunity — maybe around six months — when skilled AI users can make quick profits before the market levels out.”
The greater contribution of Sora2, Peng believes, lies in offering the industry an entirely new path to monetization.
Over the past few years, most major AI models and vertical applications have relied on a subscription-based model — charging users monthly or quarterly fees for a set amount of API usage.
But in a recent blog post, Sam Altman announced that OpenAI will introduce a new “IP revenue-sharing” mechanism. This approach has two parts:
-
More granular control for rights holders — allowing IP owners to define if and how their characters or assets can be used;
-
A profit-sharing model — when users generate videos with licensed characters and earn revenue, the copyright holders receive a share of the profits.
This marks a major conceptual breakthrough. For years, global IP giants such as Disney, Marvel, Warner Bros., and Nintendo have fought hard to protect their properties, trying to prevent unauthorized use by AI content creators. For example, Midjourney, the popular text-to-image AI app, was sued multiple times between June and September this year by Disney, Universal Pictures, and Warner Bros.
Altman’s proposal, however, turns an adversarial relationship into a win-win collaboration. For major studios and IP holders, allowing their assets to be used by creators across industries expands brand reach while creating a new source of income. For digital creators, being able to legally use iconic characters and worlds provides richer creative inspiration and helps reduce content homogenization across platforms.
Drawing on YouTube’s success, this kind of revenue-sharing model could also foster a healthier ecosystem for the AI video industry.
YouTube spent years developing a sophisticated copyright recognition and revenue-sharing system. When a creator uses a copyrighted song or movie clip in their content, YouTube’s system automatically identifies and records it, then distributes the ad revenue from that video among the creator, the platform, and the copyright owner based on pre-agreed ratios.
This win-win structure has been a major factor behind YouTube’s rise as one of the most influential video platforms in the world. Forecasts suggest that by 2025, YouTube’s global monthly active users (MAU) will surpass 2.85 billion, reaching over 51% of all internet users worldwide.
Sam Altman’s vision is to build a similar revenue-sharing ecosystem for Sora2. When users create videos using Sora2, the platform would track and log each piece of content; if the video involves a licensed IP, it would automatically allocate earnings among the creator, platform, and rights holder according to the agreed terms.
According to the China Industry Research Institute’s report “2025–2030 Global AI Video Industry Panorama and Investment Outlook,” the global AI video market was already worth $42 billion in 2023.
The logic is simple: if creators succeed, the platform succeeds. Both AI image generation and AI video creation have already gained widespread acceptance among creators. If Altman’s model takes off, the Sora app will evolve from a mere creation tool into a digital economy ecosystem — one that connects IP licensors with millions of creators worldwide, tapping directly into a multi-billion-dollar global market.
However, there’s another side to this evolution. As AI-generated videos become so realistic that they’re indistinguishable from reality, the line between truth and fabrication blurs — and “seeing is believing” may no longer hold true. Distinguishing between the real and the artificial will become a growing challenge.
One industry insider put it bluntly:
“We don’t need to obsess over what’s real or fake — learning to adapt will become the new normal.”
For instance, in livestream e-commerce, virtual hosts have already become mainstream. According to QYResearch, the global virtual idol and virtual streamer market reached $1.083 billion in sales in 2023, and is expected to grow to $5.129 billion by 2030.
“Many viewers no longer bother distinguishing whether a streamer is human or virtual,” the insider continued.
“When I make a purchase decision, it’s not based on whether the person on screen is real — it’s about their tone, attitude, and the atmosphere they create, or simply whether I trust the product’s brand in that livestream.”