OpenAI revealed this week that it is discontinuing its Sora app and associated video models, a mere six months after the app's debut. This step aligns in many ways with recent observations about the company's strategic shift toward enterprise and productivity tools, likely in preparation for a potential initial public offering. In fact, one perspective offered was that OpenAI's choice to close Sora demonstrated "a sign of maturity that was nice to see in an AI lab."

However, the end of Sora—coupled with reports that ByteDance has postponed the global launch of its Seedance 2.0 video model—may also serve as a sobering moment for developers of AI video tools and for proponents who suggest these technologies will imminently displace Hollywood. A preview of the discussion, condensed for brevity and clarity, follows.

Anthony emphasized that the issue extends beyond the app itself. He found the app particularly unappealing, describing it as a social network devoid of people, consisting of "nothing but slop." More broadly, it appears OpenAI is scaling back nearly all its video initiatives. Reporting from the Wall Street Journal indicated this shift is fundamentally about OpenAI concentrating on business, enterprise, and programming products as it considers going public. Consequently, this consumer social app and video projects more widely are not current priorities.

Sean noted he never truly used the app, as the concept was off-putting for several reasons. He suggested this episode likely served as an internal reminder for OpenAI about the role of luck in the monumental success of ChatGPT. While acknowledging the genuine value and sustained popularity of ChatGPT, Sean observed that Sora's launch carried an assumption that the company could effortlessly replicate that success, envisioning partnerships with entities like Disney. This situation starkly illustrates that there is no guaranteed shortcut to creating legendary consumer products; users must perceive meaningful value for a product to endure.

Kirsten actually commended OpenAI for this decision. She pointed out that while the "move fast and break things" ethos is often mocked, there is merit in companies that can rapidly iterate and terminate underperforming products without viewing it as a failure. Significant financial losses were involved, including a reported billion-dollar deal with Disney. Without access to internal financials, the costs and long-term value are unclear, but Kirsten argued that while the creative output was interesting, the decision to shutter the project reflected a mature approach rarely seen in AI labs.

Regarding the implications for OpenAI, Anthony stated this move is entirely consistent with the company's rumored forward strategy. It does not appear to be a major setback for the future of generative AI. The timing is notable, as it coincides with reports about ByteDance's Seedance model. Delays to Seedance 2.0 are attributed to engineering and legal hurdles, particularly concerning the integration of intellectual property protections—issues apparently not treated with sufficient gravity initially. This represents a collective reality check. Earlier hyperbolic claims, even from within Hollywood, suggested that typing prompts to create feature films was the imminent future. It is now evident that due to myriad technical and legal complexities, achieving that vision is neither easy nor close at hand.

Sean added a final point, noting this is among several decisions seemingly made since Fidji Simo assumed day-to-day operational leadership—a significant change within OpenAI. He believes that with time, as her tenure and influence over consumer product decisions become clearer, it will be easier to look back and recognize the magnitude of this transitional period for the company.

Loading the player…