OpenAI has announced ambitious plans to develop a new generation of artificial intelligence systems. These systems will be capable of performing advanced scientific research tasks, reflecting the company’s desire to expand the scope of artificial intelligence to include areas beyond just writing texts or generating images.
In a live broadcast organized by the company, CEO Sam Altman revealed that the development team is currently working on creating an AI-powered research assistant. This assistant will be at the level of an academic trainee and is expected to be completed by September 2026. He added that the company plans to launch a fully integrated AI researcher by 2028, capable of managing entire scientific projects independently.
Jakub Pachocki, Chief Scientist at OpenAI, explained that the ultimate goal of the project is to reach a system that can design scientific experiments, test hypotheses, and arrive at new discoveries without direct human intervention.
He stated in his statements that the company believes that deep learning systems may be less than a decade away from reaching the stage of “superintelligence.” That is, to become capable of outperforming humans in most mental tasks, and he added that this ambitious goal may be achieved if the pace of development and funding continues at the same current rate.
The company relies on two main strategies to achieve this goal. The first focuses on improving deep learning algorithms to become more efficient in learning and logical analysis. The second is based on increasing the computing power dedicated to the models, so that the systems are given more time and ability to think and solve complex problems.
The company believes that this combination of software development and computing power will give artificial intelligence the ability to face scientific challenges that exceed current human capabilities.
Regarding funding, Altman stated that the company’s new profit structure allows it to attract huge investments to expand its infrastructure. He pointed out that OpenAI plans to create a computing capacity of up to 30 gigawatts, at an estimated cost of about $1.4 trillion, a step he considers necessary to enable future models to function as real scientists and not just digital assistants.
The company confirmed that the goal of this project is not limited to showcasing technical capabilities, but aims to accelerate the pace of scientific research in multiple fields. For example, developing new materials, medical treatments, and improving energy efficiency.
It explained that the non-profit OpenAI Foundation will continue to oversee this scientific path through a $25 billion fund dedicated to supporting safe AI-based research.
Despite the ambitious timeline, OpenAI affirmed that it maintains a balanced approach between innovation and caution. It stressed that its new organizational structure is designed to ensure public oversight and transparency in its work.
The company stated: “We seek to build artificial intelligence that serves everyone and contributes to the advancement of humanity, not to replace it.”
