Shutting down products does not mean abandoning research: Co-founder of OpenAI states that Sora world model technology has shifted to robotics.

According to 1M AI News monitoring, OpenAI co-founder Greg Brockman revealed in the same interview that the research program behind Sora was not terminated when the product was shut down, but instead was folded into the robotics direction and continued to be pushed forward. He said that Sora’s world model and the core GPT reasoning series belong to “two different branches of the technology tree,” and that pursuing both branches at the same time—given limited compute—“is very difficult for us.” Therefore, the consumer-level application for Sora was cut, but the research team’s work shifted toward “an application that is very clearly going to be transformative: robots.”

Brockman explained why the robotics direction is still at the research stage rather than a product priority: “Robots aren’t truly mature enough yet to be deployed at scale. We’ll see a real takeoff in the knowledge work field next year—not robots.” This is consistent with OpenAI’s logic of concentrating all productization resources on the GPT reasoning series. He also emphasized that the GPT series isn’t just about text; capabilities such as voice interaction and image generation are based on different fine-tuning of the same model architecture, and do not constitute separate branches.

He also responded to comments by Demis Hassabis, the head of Google DeepMind, who previously said that image/video generation is closest to AGI. He admitted, “Absolutely possible, yes.” By choosing to bet on the GPT reasoning route, OpenAI means “you have to make a choice, take action,” and he cited an early OpenAI credo: “The sum of random vectors is zero, but if you align vectors, you can move in one direction.”