Daisuke Okanohara, Co-founder and President & CEO, Preferred Networks, Inc.

Transforming AI into an infrastructure that supports society.

Daisuke Okanohara, Co-founder and President & CEO of Preferred Networks, Inc.

■ Profile
Preferred Networks, Inc. Co-founder and CEO
Matlantis Co., Ltd. President and CEO
He holds a PhD in Information Science and Engineering. While a graduate student at the University of Tokyo, he co-founded Preferred Infrastructure with Toru Nishikawa and others. In 2014, he founded Preferred Networks (PFN) to accelerate the practical application of AI. As Chief Technology Officer, he led PFN's technology development and commercialization, including the domestically produced generative AI platform model PLaMo™ and the general-purpose atomic-level simulator Matlantis™. He became President of PFN in November 2025. He has received numerous awards and authored many books.

"Computer performance will continue to increase exponentially." Based on this conviction, Daisuke Okanohara was one of the first to foresee a future where AI would become a social infrastructure. Preferred Networks is challenging the transformation of the real world, such as robots and semiconductors, beyond the virtual world. We delve into the belief behind integrating AI into people's lives.

During my first and second years of university, I belonged to the ESS (English Speaking Society). I gained experience in things that, in retrospect, were the foundations of management, such as producing English plays and managing a retreat for over 100 people as vice president. In parallel with that, I also participated in the "Mitou Youth" project of the Information-technology Promotion Agency, Japan (IPA). This is a publicly solicited program that provides development funds to young engineers, and I was selected three times. In fact, the results of my final year led to the search engine business of Preferred Infrastructure (PFI), which I co-founded with Nishikawa and others, and which is the predecessor of my current company.
In addition, I interned at Google's headquarters in the United States. There, I spent one or two months involved in actual product development, and I was so impressed by the process that I quickly adopted it in our company after returning to Japan.
My conviction in the potential of AI stemmed from my understanding that computer performance would continue to increase exponentially. At the time, AI was still considered impractical, even referred to as "artificial unintelligence," but unlike other industries, the world of computers has seen performance improvements of 1.5 times annually for decades. In 10 or 20 years, performance would increase hundreds of millions of times. I was convinced that AI would one day become an essential part of society's infrastructure.
In 2012, the "AlexNet shock" occurred, when deep learning achieved dramatic results in the field of image recognition. My co-founder Nishikawa and I decided that we should dedicate ourselves entirely to this, and we separated our existing search engine business and other ventures to launch "Preferred Networks (PFN)," which specializes in connecting AI with the real world.

■ Four pillars for solving real-world problems

Our business is primarily built on four pillars. The first is the provision of AI-powered solutions and products. Many companies face the challenge of wanting to accelerate the speed of new product development but not being able to fully utilize their internal data. We help them dramatically accelerate this process using our own AI technology. For example, "Matlantis," a general-purpose atomic-level simulator that leads to the success of new material development; "MiseMise," which uses AI to analyze camera footage and understand and visualize the timing of stocking shelves and the condition of shelves in supermarkets; and "Kachaka," a small, highly maneuverable autonomous transport robot. Through these products, we are promoting digital transformation in the workplace.
The second pillar is semiconductor development. We are developing our own AI semiconductor, the "MN-Core series," with the aim of accelerating AI learning and the use of generative AI.
Thirdly, our business involves developing supercomputers, which are essential computing resources for AI development, and providing that computing power as a cloud service.
The fourth initiative is the construction of "large-scale language models," which form the foundation of AI intelligence. In Japanese society, which is facing a declining population, it is necessary to use AI to support rural areas and specific occupations that are suffering from labor shortages. Furthermore, we plan to focus on using the power of AI to promote Japan's strengths in manufacturing and entertainment to the world.

■ "Learn or Die": A challenge to the real world

From the very beginning, we were determined to apply AI not merely to the virtual world, but to real-world industries such as manufacturing, logistics, robotics, and materials development. Because implementing AI in the real world requires enormous time and cost, very few companies ventured into this area at the time. However, we deliberately focused on this area from the start and have steadily accumulated research.
At the core of our company is the motto "Learn or Die." When venturing into uncharted industrial fields such as automobiles, industrial equipment, and drug discovery, we value this spirit above all else. We are relentlessly learning about new areas and continuously adapting. It is this consistent attitude that has earned us the trust of global companies such as Toyota Motor Corporation and FANUC, and has led to them choosing us as their joint development partner.
Looking ahead, we aim to elevate AI to the level of "social infrastructure," like electricity and water. Just as Japan once dominated the world in manufacturing, we intend to create a future where we can export next-generation AI products originating in Japan to the entire world.

■ "Curiosity" and "Proactiveness" will create the future

When I want to work with someone, I look for someone who not only has expertise, but even more so, someone who values ​​teamwork and treats others with respect. And "proactiveness" and "curiosity" are also important.
The world of technology is constantly changing. Even today's expertise will become outdated in a few years. That's why we provide a learning environment, and we want you to have a "desire to learn" and be able to absorb new things with interest. We value how each individual can grow and act proactively.

■Message to students

While you're a student, don't overthink whether an experience will be useful in the future; just try out a variety of experiences. Once you enter the workforce, you tend to focus on just one job, but the diverse experiences you gain as a student will connect in unexpected ways later on. I also want you to find something that "others don't want to do, but you're the only one who loves it." From a supply and demand perspective, places where everyone wants to do are highly competitive, but when your unique "passion" perfectly aligns with a problem or need in the world, it becomes an irreplaceable and powerful weapon. Trust your curiosity and forge ahead.

Interview with Ayami Matsuyama, a first-year student at Aoyama Gakuin University, published by Student Newspaper Online on November 13, 2026

Chiharu Yamada, 2nd year student at the University of Information Management and Innovation / Ayami Matsuyama, 1st year student at Aoyama Gakuin University