OpenAI says deep learning systems are advancing rapidly, with models increasingly capable of solving complex tasks faster. So fast, in fact, that internally, OpenAI is targeting achieving an in-house research assistant by September 2026 and a fully automated “legal AI investigator” by 2028, CEO Sam Altman said during a livestream Tuesday.
The ambitious timeline comes on the same day that OpenAI finalized its transition to a for-profit corporate structure, shedding its non-profit roots. This restructuring frees OpenAI from the constraints associated with its non-profit charter, while also opening up new opportunities for fundraising.
Jakub Pachocki, OpenAI’s chief scientist, joined Altman on the live stream. He described this AI researcher — not to be confused with a human doing AI research — as “a system capable of autonomously contributing to larger research projects.”
“We think it’s possible that deep learning systems are less than a decade away from superintelligence,” Pachocki added. He described superintelligence as systems smarter than humans in a large number of critical actions.
To achieve these goals, OpenAI is betting on two key strategies: continued algorithmic innovation and dramatic scaling of “test time computation” — essentially how much time models spend thinking about problems. Current models can handle tasks with a time horizon of about five hours and match top humans in competitions like the International Mathematical Olympiad, Pachocki said. But he thinks that horizon will expand rapidly, in part by allowing models to spend much more computing resources thinking about complex problems. For major scientific discoveries, he said, it would be worth devoting the computing power of entire data centers to a single problem.
OpenAI says these goals are consistent with the company’s overall push to advance scientific research and enable AI to make discoveries faster than human researchers, tackle complex problems beyond current human capabilities, and dramatically accelerate technological innovation in many areas such as medicine, physics and technological development.
Altman also said the restructuring creates a framework to support OpenAI’s aggressive timeline for AI research assistants while maintaining a commitment to responsible AI development. Under the new structure, the non-profit OpenAI Foundation, which focuses on scientific progress, will own 26% of the for-profit and govern the research direction. The nonprofit also has a $25 billion commitment to use AI to treat disease and will help manage AI research and safety initiatives.
Techcrunch event
San Francisco
|
27-29 October 2025
According to Altman, the for-profit arm’s ability to raise more capital means it can scale up the necessary infrastructure construction to achieve scientific advances. Altman said OpenAI has committed to 30 gigawatts of infrastructure, a $1.4 trillion financial commitment, over the next few years.
