OpenAI’s plans extend beyond chatbots and image generation. In a wide-ranging interview at a StrictlyVC event, co-founder and CEO Sam Altman discussed a video model in development, the company’s relationship with Microsoft, and the questions AI raises for education and safety.
A video model is in the works
Alongside ChatGPT and the digital art generator DALL-E, OpenAI is developing a video model, Altman confirmed. He did not offer a launch schedule, describing it as a legitimate research project whose arrival could be soon or could take a while.
That uncertainty is a useful distinction: the project exists, but its release timing is not settled. Altman said he would not make a competent prediction about when it would be ready. His comments indicate that OpenAI is exploring another kind of generated media, while leaving open how long that work will take.
Altman also clarified that OpenAI can create its own software and services as well as license its technology to other companies. Its partnership with Microsoft, which first invested in OpenAI in 2019, is not exclusive. The company therefore has room to pursue products itself while continuing to work with outside partners.
GPT-4 will come when OpenAI considers it safe
Asked about GPT-4, the language model behind ChatGPT, Altman gave no release date. He said it would come out when OpenAI was confident it could release it safely and responsibly. That answer put readiness, rather than a calendar commitment, at the center of the launch decision.
He also urged people not to assume GPT-4 would amount to artificial general intelligence, or AGI. Altman said OpenAI does not have an actual AGI, and warned that people expecting GPT-4 to meet that description would be disappointed.
In his view, the transition toward AGI may be less like a sudden arrival and more like a gradual, blurry shift. As AI capabilities approach that threshold, he said, it becomes harder to give a precise answer about when it will happen. The uncertainty itself, in his account, reflects how difficult it is to draw a clear boundary.
Safety and education are already in the conversation
AI safety was a central topic in the interview. Altman described OpenAI’s internal efforts to test systems and study their impacts, along with work involving external auditors, external red teamers, other labs and safety organizations.
Those measures exist alongside public concerns about tools already released. Some educators are blocking access to ChatGPT because they worry students will use it to cheat. Altman acknowledged that generative AI is affecting education and academic integrity, and said society needs to begin working out how to live with those changes.
He argued that introducing products while the stakes are relatively low gives society time to adapt before the technology becomes more widespread. That position does not mean every concern is resolved; it frames early releases as an opportunity to learn about their effects while there is still time to respond.
Detection tools may help, but educators will still adapt
Altman said OpenAI and other AI companies would experiment with watermarking and other verification methods to help identify AI-generated writing. But he cautioned against treating detection as a lasting solution: a determined person may be able to get around those methods, he said.
He compared the educational adjustment to earlier technology shifts. Calculators changed what math classes test, while Google made memorizing facts less important. Altman described deep learning models as a more extreme version of those changes, with benefits that could also be more substantial.
Teachers, he said, have expressed both anxiety about homework and excitement about AI as a personal tutor for each child. That contrast points to the practical challenge: schools may have to reconsider assignments and learning goals while exploring useful ways students can work with these systems.
Altman’s remarks leave major questions open, including when OpenAI’s video model or GPT-4 will arrive and how education will adjust. They do, however, offer a picture of the company’s stated direction: continued product development, safety work, and an expectation that people and institutions will need to adapt as AI tools become part of everyday life.