OpenAI plans to launch a "Desktop SuperApp" that deeply integrates chat, programming, and browser functions.

According to informed sources, OpenAI is developing a desktop application that aims to integrate the core logic of ChatGPT chat interface, Codex programming environment, and Atlas web browser in order to compete with rivals such as Anthropic PBC and maintain user engagement. The sources revealed that this single application will integrate ChatGPT, Codex, and the AI web browser ChatGPT Atlas that was launched last year. It is currently unclear when this application will be released, but OpenAI will continue to provide the standalone ChatGPT application. Specifically, this product, codenamed "Desktop Super App," is expected to deeply integrate the previously released Operator automation agent, enabling it to not only have text generation capabilities but also utilize screen-aware technology to perform complex tasks on the computer interface. Additionally, by embedding the underlying architecture of the Atlas browser into the application, AI models can more accurately parse webpage structures, enabling a seamless transition from information retrieval to automated form filling, data extraction, and online booking. Furthermore, for the developer community, the application will optimize code writing and real-time execution synchronization, aiming to provide an immersive programming space without the need for frequent switching between an IDE and an AI window. The sources mentioned that Fidji Simo, Chief of OpenAI's Applications Business, will lead the integration of the products into a single application and oversee marketing efforts, while Greg Brockman, President of OpenAI, will collaborate with Simo in advancing the product development. OpenAI declined to comment on this. From the perspective of market competition and business landscape, this move by OpenAI signifies a complete restructuring of its productivity tool logic. By driving models with deep reasoning capabilities such as the o3 series, this desktop application will no longer merely assist users but evolve into a task executor with system-level permissions. Currently, OpenAI has gradually opened up features such as "Work with Apps" in the existing desktop versions, allowing AI to read and understand the content of other applications with authorization. It is worth mentioning that in recent years, OpenAI, Anthropic, Google under Alphabet, and other developers have released multiple AI models that can handle increasingly complex tasks - from analyzing company financial reports and writing code to generating realistic and mind-blowing images and videos.