Jeff Dean, one of the longest-serving and most influential executives at Google, is stepping down from the search giant to launch his own AI startup. Coming with him as co-founders are several other top researchers at the company, including Sanjay Ghemawat, a top engineer and senior fellow at Google; Quoc Le, a key AI researcher and founding member of Google Brain; and Oriol Vinyals, a senior research scientist at Google DeepMind. Dean reportedly plans to serve as CEO.
Together, the group is starting Discovery Loop, a public benefit corporation that seeks to use AI to turbo-charge scientific research. Discovery Loop says it plans to use high-octane algorithms to initiate and iterate thousands of experiments simultaneously, with the goal of partially automating the research process and expanding the scale at which experimentation can be conducted. The startup is also interested in using AI to help create more powerful AI (a process known as recursive self-improvement), which would cut human iteration out of the loop entirely.
“While science and engineering have tremendously advanced society over past centuries, progress has traditionally relied on slow, sequential human iterations, creating a significant bottleneck,” the company said in a press release. “Discovery Loop is developing advanced AI systems that leverage massive computational scale to fundamentally transform the speed and efficiency of innovation by automating complete experimental loops.”
Using AI to accelerate scientific discovery has been a key interest of the science and tech communities for many years, but until recently, it remained a largely experimental field with limited commercial application. The new venture aims to change that by building a platform that can run experiments at a scale and speed that no human team could match. Instead of a scientist manually setting up a hypothesis, designing an experiment, collecting data, and analyzing results, Discovery Loop envisions a future where AI systems can do all of that with minimal human oversight.
“The next great frontier for AI is to go beyond answering questions and to begin making discoveries,” the founding team said in a joint statement. “By fundamentally accelerating how engineering and scientific discovery are conducted, we can deliver the benefits of transformative technologies to the world far sooner.”
A Legendary Career at Google
Dean has worked at Google since 1999 and was Google’s 30th employee. Over the years, he has contributed to Google search’s core infrastructure, including its crawling and indexing system and its query-serving system. He has also had a significant impact on Google Gemini’s multimodal models and played a leading role in the company’s early AI research. His work helped lay the foundation for many of the technologies that have made Google one of the most powerful companies in the world.
Dean’s influence extends far beyond the products he helped build. He is widely regarded as one of the most respected engineers in the tech industry, known for his ability to tackle complex problems and inspire those around him. He has co-authored dozens of influential papers on distributed systems, machine learning, and large-scale data processing. His contributions to the open-source community, including the MapReduce framework and the TensorFlow machine learning library, have shaped the way modern AI is developed and deployed.
His decision to leave Google marks the end of an era for the company, where he has been a central figure for more than two decades. Dean’s departure is particularly significant because he remains one of the few executives with deep technical expertise who also holds considerable strategic influence. His move to found Discovery Loop is seen as a vote of confidence in the potential of AI-driven scientific discovery, a field that is attracting increasing attention from both researchers and investors.
The Co-Founders
Sanjay Ghemawat, who is joining Dean as a co-founder, is a senior fellow at Google and one of the company’s most accomplished engineers. He has worked alongside Dean for many years, and together they developed key technologies that underpin Google’s infrastructure. Ghemawat is known for his work on distributed systems and large-scale data processing, and he co-created MapReduce with Dean in the mid-2000s. That framework became a cornerstone of modern big data analytics and was widely adopted across the tech industry.
Quoc Le is another prominent co-founder. As a founding member of Google Brain, he played a pivotal role in advancing deep learning research at the company. His work on unsupervised learning and neural networks has been instrumental in many of the AI breakthroughs that have emerged from Google over the past decade. Le’s research has focused on building systems that can learn from vast amounts of data with minimal supervision, a goal that aligns closely with Discovery Loop’s mission to automate the research process.
Oriol Vinyals, a senior research scientist at Google DeepMind, rounds out the founding team. Vinyals is best known for his contributions to sequence-to-sequence learning, which has become a fundamental technique in natural language processing and machine translation. He has also worked on reinforcement learning and meta-learning, areas that could be crucial for developing AI systems capable of designing and running their own experiments. Together, the four co-founders bring a depth of expertise that covers everything from infrastructure to algorithmic innovation.
Discovery Loop’s Mission
Discovery Loop is being established as a public benefit corporation, meaning it is legally required to consider the impact of its actions on society and the environment, not just on shareholders. This structure reflects the founders’ stated goal of using AI for the public good, particularly by accelerating progress in fields like medicine, chemistry, and materials science. The company is not just focused on commercial applications; it aims to make scientific research itself more efficient and accessible.
The core idea behind Discovery Loop is to close the loop between hypothesis formation, experimentation, and analysis. In traditional research, scientists must manually design each experiment, which is time-consuming and limits the number of experiments that can be conducted. Discovery Loop plans to use AI to automatically generate hypotheses, design experiments, run them in simulation or in the physical world, and then learn from the results. This approach could dramatically increase the scale at which research is conducted, enabling thousands or even millions of experiments to be carried out simultaneously.
One of the most ambitious aspects of Discovery Loop is its interest in recursive self-improvement. This concept involves using AI to create more capable AI, potentially creating a positive feedback loop where each generation of AI is designed by the previous generation. While this idea has been discussed in theoretical terms for decades, it has rarely been pursued as a concrete engineering goal. Discovery Loop’s founders believe that automating the research process could eventually lead to AI systems that are capable of making breakthroughs in AI itself, accelerating the pace of innovation even further.
Financial Backing and Market Context
The company has received financial support from a number of different sources, including Google’s parent company Alphabet. The initial funding round is being co-led by Radical Ventures and Khosla Ventures, the company announced Wednesday. Kleiner Perkins, Lightspeed, and Doerr Capital also participated. This syndicate includes some of the most prominent venture capital firms in the technology sector, reflecting strong confidence in Discovery Loop’s vision and leadership team.
The involvement of Alphabet is notable, as it suggests that Google’s parent company is willing to support ventures that could eventually become competitors or disruptors in the AI space. This is not unprecedented; Alphabet has previously backed spin-offs and independent research initiatives that operate outside its core businesses. However, it does indicate that the relationship between Dean and Google remains amicable, at least publicly, despite his departure from an executive role.
Discovery Loop is entering a market where there is already significant interest in using AI for scientific discovery. Companies like DeepMind have made headlines with breakthroughs in protein structure prediction, while other startups are using AI to identify new materials and drug candidates. What sets Discovery Loop apart is its focus on automating not just one part of the research process but the entire experimental loop. This ambitious approach could produce results that are impossible to achieve with current methods.
“We think there is opportunity for AI to more fully automate what has traditionally been a very human-intensive experimental loop,” Dean told the New York Times. “You will get both a higher quantity and a higher quality of experiments, and that will lead to scientific breakthroughs and advances.”
The creation of Discovery Loop also raises important questions about the future of AI research and the role of large tech companies. As top researchers leave established giants like Google to start their own ventures, we may see a shift in where cutting-edge AI innovation takes place. These new startups often have more flexibility and can pursue riskier research directions that might not fit within the strategic priorities of a large corporation. That could lead to a proliferation of AI-driven scientific discovery efforts, each focused on different problem domains.
For now, Discovery Loop is still in its early stages. The company has not publicly disclosed detailed plans for its first products or research programs, but its founders have stated that they intend to build something that can have a tangible impact on the speed of scientific progress. With a world-class founding team and substantial financial backing, Discovery Loop is well-positioned to become a major player in the emerging field of AI-powered research.
Jeff Dean’s departure from Google is a milestone for the AI community, and his new venture will be closely watched by researchers, investors, and tech enthusiasts alike. The success or failure of Discovery Loop could help determine whether the idea of fully automated scientific discovery is a practical reality or just an ambitious dream. Either way, this startup marks a new chapter for some of the most influential minds in artificial intelligence.
Source: TechCrunch News