Documents released during the antitrust court proceedings revealed OpenAI’s plans for the further development of ChatGPT. According to the company’s strategy for the first half of the next year, ChatGPT is set to transform into a “super-assistant” that deeply understands the user, considers their interests, and assists with any tasks that a smart and reliable person with computer access could perform.
The document emphasizes that current models, such as 02 and 03, are already advanced enough to perform complex agent tasks. The application of features like computer usage will allow ChatGPT not only to answer questions but also to act on behalf of the user. Additional capabilities, including multimodality and a generative interface, will ensure convenient interaction and a wider range of use scenarios.
OpenAI plans for the “super-assistant” to help with everyday tasks — from vacation planning, finding housing or a lawyer to organizing a calendar, sending emails, and creating task lists. There is also mention of support for more specialized tasks, such as programming.
The company emphasizes its aspiration to make ChatGPT accessible in various areas of life: at home — for answering questions, playing music, or selecting recipes; on the go — for finding routes and venues; at work — for taking notes and preparing presentations. At the same time, OpenAI acknowledges that the growing popularity of the model creates additional infrastructure demands, prompting the company to focus on building data centers.
