About Me

I lead Worldsim as a VP of Research at OpenAI. I’m currently working on bootstrapping our new robotics effort to bring the intelligence of our video generation models to the physical world.

Here are some of the projects I’ve worked on:

• Sora 2 (October 2025): Our team advanced the state of the art in physical realism and prompt fidelity for complex scenes. We released our model in the Sora app. My work focused on controllability and language grounding – helping users turn vague ideas into compelling stories and giving power users precise control over cast and scene dynamics.
• Sora (December 2024): I built the team that created the model and the product experience at sora.com. Our model advanced the state of the art in generating realistic videos up to a minute long from user prompts.
• DALL·E 3 (September 2023): I led the team that created the model and deployed it in ChatGPT. Our model advanced the state of the art in faithfully interpreting complex user prompts, including text rendering – a big step toward making image generation practically useful.
• DALL·E 2 (April 2022): I co-created the first text-to-image generation model to be deployed in production, introducing inpainting, variations, and blending – features now widely used in image generation products. You can read more about how it works here.
• DALL·E (January 2021): I created the first text-to-image generation model with the ability to combine a wide range of concepts from language in plausible ways, a step toward machines being able to manipulate visual concepts with human-like flexibility.

Find me on Twitter here or reach me by email at _@adityaramesh.com.