OpenAI Sets Fully Automated AI Researcher as 'North Star' Goal - Intern-Level by September, Full Version by 2028

According to MIT Technology Review, OpenAI has designated the development of a fully automated AI researcher as its top internal priority, dubbed the 'North Star.' The initiative is led by Chief Scientist Jakub Pachocki, who envisions building AI systems capable of conducting research with the coherence and persistence of human experts. The concrete milestones include developing an 'autonomous AI research intern' by September 2026 and achieving a complete multi-agent research system by 2028. This AI research intern is expected to autonomously formulate hypotheses, design and execute experiments, analyze results, and write papers on given research topics. The strategic intent behind this goal is to accelerate AI research itself. If AI can conduct AI research, the pace of technological progress could accelerate exponentially, potentially realizing the concept of an 'intelligence explosion.' Pachocki has identified overcoming current AI limitations in 'sustained attention' and 'long-term planning' as the greatest technical challenges. While human researchers can maintain coherent research programs over weeks or months, current AI systems lack such sustained cognitive capabilities. This announcement is expected to intensify the industry-wide race for 'AI-driven AI development.' Google DeepMind and Anthropic are pursuing similar efforts, and the realization of fully automated AI researchers is positioned as a critical step toward AGI (Artificial General Intelligence). Meanwhile, debates are intensifying about the impact of research automation on human researcher employment and the risks of uncontrollable AI self-improvement cycles.