Will DePue, a key engineer behind OpenAI’s groundbreaking text-to-video model Sora, has announced he is rejoining the company to lead a new team focused on a bold, high-risk pursuit: artificial superintelligence. In a post on X, formerly Twitter, DePue revealed he is returning to OpenAI after a period away to form a small, tightly-knit team with Troy Luhman and Eric Luhman—both core members of the Sora development group. DePue described the initiative as an “incredibly high-risk bet” with a “small, but significant, chance of leading to ASI,” or artificial superintelligence. He emphasized the team’s focus on uncharted territory, stating they are “keeping the team tight but we’re open to high-slope researchers & engineers interested in making progress in uncharted territory.” At press time, OpenAI had not issued any official statement about the new team, and the company did not respond to a request for comment from Business Insider. Details about the team’s goals, structure, or technical approach remain scarce. DePue, born in 2003, studied computer science at the University of Michigan before taking a leave of absence, which he described on his personal website as a “drop out” period. He first joined OpenAI in July 2023, where he contributed to training early versions of ChatGPT and played a central role in developing Sora. His LinkedIn profile confirms his return to the company. During his time away, DePue worked on several personal projects, including a browser-based GPT implementation, a Turing-complete computer built entirely within Figma, and an ambitious dataset project called the “Alexandria Index,” aimed at organizing and indexing vast amounts of AI-relevant information. While little is publicly known about Troy and Eric Luhman, both have co-authored machine learning research papers, suggesting strong technical credentials. Together, the trio formed the core team behind Sora, which was first unveiled in February 2024 and officially launched in December. A follow-up version, Sora 2, was released in September. OpenAI has long stated that achieving artificial general intelligence—AI capable of human-level reasoning—is its ultimate mission. However, in a January blog post, CEO Sam Altman signaled a shift toward even more ambitious goals, explicitly mentioning superintelligence. He wrote that superintelligent systems could “massively accelerate scientific discovery and innovation well beyond what we are capable of doing on our own,” potentially unlocking unprecedented levels of abundance and prosperity.