DMR News

Advancing Digital Conversations

Tokyo-based startup Sakana AI secures $30 million funding to innovate with streamlined AI models

ByYasmeeta Oon

Jan 17, 2024

Tokyo-based startup Sakana AI secures $30 million funding to innovate with streamlined AI models

Sakana AI, a groundbreaking artificial intelligence startup headquartered in Tokyo, was co-founded by two distinguished former Google engineers. The company has recently made headlines by securing a remarkable $30 million in seed funding. This impressive financial backing comes from a collection of esteemed technology investors, highlighting the significant interest in Sakana AI’s innovative approach to AI development.

Founded just a year ago, Sakana AI is already charting a unique course in the AI landscape. The startup is committed to developing AI models that are not only smaller but also more efficient, drawing inspiration from the natural world. This approach stands in stark contrast to the prevailing trend in AI, where the focus has often been on creating larger, more data-intensive models.

Leading the charge in Sakana AI’s funding round was Lux Capital, a venture capital firm with a history of investing in trailblazing AI companies such as Hugging Face. Joining Lux Capital in this venture was Khosla Ventures, a firm known for its early investment in OpenAI back in 2019. The round also saw participation from Japanese technology giants Sony, NTT, and KDDI. Their involvement is particularly noteworthy, as it represents a significant vote of confidence in Sakana AI from major players in the domestic tech industry.

The name “Sakana,” which translates to “fish” in Japanese, is a nod to the company’s nature-inspired ethos. Sakana AI looks to the collective behaviors of animal groups, such as schools of fish and flocks of birds, for inspiration. The startup’s philosophy is centered around the idea that smaller, more efficiently coordinated AI models can outperform the capabilities of a single, larger model. This is a departure from the norm in the AI field, where the emphasis has typically been on training massive models using extensive datasets.

Sakana AI’s co-founders, David Ha and Llion Jones, bring a wealth of experience from their time at Google, where they led AI research groups. Jones, in particular, is known for co-authoring a landmark paper on the Transformer model in 2017. This model is a cornerstone of modern AI, underpinning chatbots like OpenAI’s ChatGPT.

At a time when the AI field is heavily focused on scaling up models to achieve benchmark-beating performance, Sakana AI’s approach is refreshingly different. While the trend of building larger models has yielded impressive results, it has also sparked concerns about the environmental impact and the vast computing resources required for training and running these colossal AI systems.

Sakana AI’s executives argue that as models increase in size, they tend to become less efficient. In contrast, smaller, specialized models can

collaborate and potentially match the capabilities of larger ones. This concept is similar to how in human teams, individuals with distinct skills often outperform a single person with a broad range of abilities in complex tasks.

The founding team’s strong background and their alternative vision for AI development have attracted significant attention from both top Silicon Valley investors and prominent Japanese firms. This backing is a testament to the potential of Sakana AI’s approach and has enabled the company to establish a robust lab in Tokyo after only a year of independent operation.

With this new influx of capital and partnerships with companies like NTT, Sakana AI is well-positioned to expand its team and further refine its nature-inspired AI techniques. The early-stage backing from both U.S. and Japanese tech giants is a clear signal of confidence in Sakana AI’s potential to pioneer a new paradigm in AI from Asia. This development is particularly significant given the current dominance of the U.S. and China in strategic AI technologies.

Sakana AI’s philosophy of smaller, collaborative AI models is not just an alternative approach to AI development; it’s a response to the growing concerns about the sustainability and practicality of current AI practices. The company’s focus on efficiency and nature-inspired solutions is poised to bring a fresh perspective to the AI field, which has long been dominated by a race to build ever-larger models.

In addition to its unique approach to AI, Sakana AI’s success in securing funding from both domestic and international investors highlights the growing interest in diverse and innovative AI strategies. This interest extends beyond traditional tech hubs, indicating a shift in the global AI landscape. As AI continues to evolve, Sakana AI’s approach could lead to more sustainable, efficient, and

creative applications of artificial intelligence, potentially reshaping the industry.

The company’s innovative strategy also reflects a broader trend in technology towards more environmentally conscious and sustainable practices. By focusing on smaller, more efficient models, Sakana AI is addressing the significant energy demands and carbon footprints associated with large-scale AI operations. This approach not only benefits the technology sector but also contributes to global efforts to combat climate change.

Sakana AI’s vision extends beyond just technological innovation; it represents a philosophical shift in the way we think about and implement AI. By drawing inspiration from nature and focusing on collaboration and efficiency, Sakana AI is exploring new frontiers in AI that could lead to more organic and adaptable systems. These systems could potentially mimic the resilience and flexibility found in natural ecosystems, offering solutions that are more in tune with human needs and environmental constraints.

Furthermore, Sakana AI’s success demonstrates the growing importance of interdisciplinary approaches in technology. By integrating insights from biology, ecology, and other fields, the company is breaking down traditional silos and fostering a more holistic understanding of intelligence, both artificial and natural. This interdisciplinary approach could lead to breakthroughs that transcend the current limitations of AI, opening up new possibilities for solving complex global challenges.

The company’s journey also highlights the crucial role of international collaboration in advancing technology. Sakana AI’s ability to attract funding and support from both Japanese and American investors underscores the global nature of technological innovation. This collaboration across borders enriches the tech ecosystem, fostering a diverse range of perspectives and expertise that can drive more robust and innovative solutions.

Looking ahead, Sakana AI’s trajectory is set to influence not just the field of AI, but also the broader technology sector. As the company continues to grow and develop its unique approach to AI, it will likely inspire other startups and established firms to explore alternative, nature-inspired strategies. This could lead to a more diverse and dynamic tech landscape, with a range of approaches that prioritize efficiency, sustainability, and collaboration.

In conclusion, Sakana AI’s recent success in securing significant seed funding is a testament to the potential of its innovative, nature-inspired approach to AI. The company’s focus on smaller, more efficient models not only sets it apart in a field dominated by large-scale operations but also aligns with broader trends towards sustainability and interdisciplinary collaboration. As Sakana AI continues to develop and refine its techniques, it stands poised to make a meaningful impact on the AI industry and beyond, paving the way for a more sustainable and collaborative future in technology.

Yasmeeta Oon

Just a girl trying to break into the world of journalism, constantly on the hunt for the next big story to share.