A seismic shift is underway in the AI research world as Jeff Dean, Google's longest-serving executive and a pivotal figure in its AI ascent, departs to found a new venture. He is joined by three other prominent researchers from the search giant, signaling a notable brain drain from one of the industry's most influential labs. The new company, named Discovery Loop, aims to revolutionize scientific discovery by deploying AI to run experiments at an unprecedented scale. Dean will serve as CEO of Discovery Loop, a public benefit corporation designed to leverage AI for accelerating scientific research. Co-founders include Sanjay Ghemawat, a senior fellow and top engineer at Google; Quoc Le, a founding member of Google Brain; and Oriol Vinyals, a senior research scientist at Google DeepMind. Their collective expertise spans core infrastructure, deep learning, and multimodal AI, giving the startup a formidable technical foundation. Main Developments Discovery Loop's mission is to use high-octane algorithms to initiate and iterate thousands of experiments simultaneously. The goal is to partially automate the research process, expanding the scale at which experimentation can occur. The company also plans to explore recursive self-improvement—using AI to create more powerful AI—which could remove human iteration from the loop entirely. Read also: Reddit Rethinks Karma Requirements With New AI Moderation Tools The startup's approach addresses a fundamental bottleneck in scientific progress: the slow, sequential nature of human experimentation. By automating complete experimental loops, Discovery Loop aims to transform the speed and efficiency of innovation, according to the company's press release. The team believes this will lead to both a higher quantity and quality of experiments, ultimately driving scientific breakthroughs. Financial backing for Discovery Loop comes from several sources, including Alphabet, Google's parent company. The initial funding round, announced Wednesday, is co-led by Radical Ventures and Khosla Ventures, with participation from Kleiner Perkins, Lightspeed, and Doerr Capital. This diverse investor group underscores the commercial potential seen in AI-driven scientific discovery. Background Dean's departure marks the end of an era at Google, where he has worked since 1999 as the company's 30th employee. He contributed to the core infrastructure of Google Search, including its crawling, indexing, and query-serving systems. More recently, he played a leading role in the development of Google's Gemini multimodal models and was a driving force in the company's early AI research. The idea of using AI to accelerate scientific discovery is not new; it has been a key interest in the science and tech communities for years. However, until recently, it remained largely experimental with limited commercial application. Discovery Loop's formation, with significant funding and top-tier talent, signals a shift toward practical, large-scale implementation of this concept. The founding team's joint statement emphasizes a frontier beyond answering questions: making discoveries. By accelerating engineering and scientific discovery, they aim to deliver transformative technologies to the world sooner. This vision aligns with a broader industry trend of applying AI to complex, real-world problems. Why It Matters The exodus of Dean and his colleagues from Google represents a significant loss for the tech giant, which has long relied on its AI expertise to maintain a competitive edge. Their departure could weaken Google's position in the AI race, especially as rivals like OpenAI and Microsoft continue to push boundaries. For the industry, it signals that top researchers are increasingly willing to strike out on their own, potentially fragmenting AI talent. For the broader scientific community, Discovery Loop's approach could dramatically shorten the timeline for breakthroughs in fields like medicine, materials science, and energy. Automating experimental loops could allow researchers to test thousands of hypotheses simultaneously, a capability that was previously unimaginable. The involvement of Alphabet as an investor also suggests that even Google's parent company sees value in this new venture, potentially hedging its bets. What's Next Discovery Loop has announced its initial funding round, but details on the company's immediate operational plans remain scarce. The team is likely to focus on building its platform and attracting additional talent in the coming months. As a public benefit corporation, it will also need to balance profit motives with its stated mission of advancing scientific discovery. The startup's success will depend on its ability to deliver on the promise of automated experimentation, a goal that has proven challenging in the past. Observers will be watching closely to see how Discovery Loop differentiates itself from existing AI research efforts, both within Google and at other tech companies