OpenAI said its internal research teams now use 3.1 agent-workdays for every human research workday, reporting it had achieved an “automated research intern” capable of carrying out defined
OpenAI says it has achieved an “automated research intern”: a supervised system that completes well-defined tasks, including work that might take a skilled researcher several days. By mid-August, its research organization ran 3.1 agent-workdays for every human workday, counting parallel agents and subagents. The figure measures runtime—not a 3.1× productivity gain—and more than half of successful four- to eight-hour tasks still required human intervention.