OpenAI CEO Sam Altman announced during a livestream on Tuesday that the company’s AI systems are evolving rapidly, with the goal of developing an “intern-level research assistant” by 2026 and a fully autonomous “AI researcher” by 2028. This comes as OpenAI transitions into a public benefit corporation, moving away from its non-profit roots to accelerate funding and innovation.
Altman explained that the structural change allows OpenAI to overcome previous limitations tied to its non-profit model, enabling the company to raise capital for large-scale infrastructure and computational needs. Jakub Pachocki, OpenAI’s chief scientist, noted that the envisioned AI researcher will be capable of autonomously conducting large research projects, rather than simply imitating human researchers.
Read More: Pakistani researcher achieves milestone in artificial vision in AI era
Pachocki emphasized that deep learning systems could be less than a decade away from achieving “superintelligence,” or intelligence surpassing humans across multiple domains. He added that OpenAI’s next phase of innovation relies on two strategies: algorithmic advancement and increased “test time compute,” meaning the models will spend longer periods solving complex problems — potentially using entire data centers for a single research task.
OpenAI’s broader mission, according to Altman, is to accelerate scientific discovery and innovation in fields like medicine, physics, and technology through AI. The company’s non-profit foundation will continue to oversee research directions, holding a 26% stake in the for-profit entity. Altman also revealed that OpenAI has committed to building 30 gigawatts of computing infrastructure — a $1.4 trillion investment — to support its long-term AI goals.