Nurturing Innovation: CIC Tokyo’s Flourishing AI Community

At the heart of Tokyo’s bustling Toranomon Hills lies a hub of innovation and collaboration—CIC Tokyo. Within its walls, a vibrant community of researchers, entrepreneurs, and technologists converge to push the boundaries of what’s possible. Among the numerous initiatives driving this spirit of innovation forward is a burgeoning AI community, which includes everyone from world-renowned researchers, young talent just embarking on a career in AI, to enthusiasts interested in understanding the impact AI will have on the world. 

To foster collaboration amongst the growing AI community, CIC Tokyo is kicking off a series of engaging events and initiatives that bring together diverse perspectives and expertise to tackle some of the most pressing challenges in artificial intelligence. 

Introducing the AI Research Tech Talk Series

One such initiative is the AI Research Tech Talk Series, a new event series created in partnership with CIC Tokyo client, Sakana AI, an upcoming world-class AI research lab. 

“At Sakana AI we have never organized a public talk before, and one of my motivations for starting this series is to build off the hype around AI right now and create an opportunity to talk about the innovation side of AI versus the surface level discussions you hear everywhere,” said David Ha, Co-Founder of Sakana AI. 

The inaugural event, which drew more than 500 attendees in person and virtually, focused on cutting edge AI development, and provided a platform for luminaries in the field to discuss their work on a range of topics. The discussions ranged from adaptive neural networks that “grow” in real-time, to brain-controlled robotic interfaces, and Sakana AI’s latest development on Evolutionary Model Merge techniques. 

Highlights from the inaugural event

  • The Sakana AI team discussed their Evolutionary Model Merge techniques which aims to automate the development of foundation AI models. This approach utilizes evolution to combine existing open source models, creating new powerful foundation models with user-specified abilities. The goal is to create a machine that can generate AI models, leveraging the collective intelligence of existing models. This approach not only addresses the resource-intensive nature of training foundation models but also shows a path to innovation in Japan by fostering fresh ideas and incremental progress in the AI field.
  • Joachim Winther Pedersen provided an overview of the work his team is doing on the Grow-AI research project, led by Professor Sebastian Risi at the IT University of Copenhagen, funded by the European Research Council. The project’s mission is to develop algorithms for growing new networks that learn and adapt in changing environments, departing from traditional fixed network structures. This approach aims to leverage compute power effectively, ensuring longevity and adaptability of AI algorithms in the face of advancing computational capabilities.
  • Kai Arulkumaran, a research team leader at Araya, discussed his work exploring the intersection between AI, neurotech, and robotics, with a focus on understanding the nature of intelligence and developing general-purpose intelligent agents. Arulkumaran discussed the challenges posed by Moravec’s paradox, where tasks seemingly simple for humans, like walking, are difficult for machines, while tasks traditionally considered complex, like playing chess or Go, have been successfully accomplished by AI. The research presented involves deep reinforcement learning, combining deep learning and reinforcement learning to create more adaptable and capable AI systems. 

Arulkumaran also highlighted his team’s ongoing work in brain-machine interfaces for various applications, including speech and facial avatar animation, assistive robotics, and controlling robots through thought signals. The goal is to leverage these technologies to improve the quality of life for individuals with disabilities and to advance human-robot interaction.

Join the AI Community 

The goal of the AI Research Tech Talk Series at CIC Tokyo is to provide a platform for thought leaders, researchers, and enthusiasts to exchange ideas, collaborate on innovative projects, and envision the future of AI technology. 

Whether you’re an AI enthusiast, a seasoned researcher, or simply curious about the future of technology, stay tuned for the next event in this series, which you can join in person or virtually with CIC Tokyo. 

The Ecosystem Japan Asia Tokyo