top of page

An Insightful Discussion with NVIDIA's CEO Jensen Huang: Exploring the Future of AI in Singapore

What an exhilarating experience it was at the media roundtable with Jensen Huang, the visionary founder and CEO of NVIDIA! Invited by Jacqueline Poh and Prime Minister Lee Hsien Loong, Huang's presence in Singapore was impeccably timed, coming just days after the unveiling of the National AI Strategy 2.0 (NAIS 2.0) – and let's face it, that's a pretty neat abbreviation!

The roundtable transcended a mere discussion. It was a profound exploration into the realms of AI, particularly in the context of Singapore's ambitious AI strategy. With Huang at the helm, we delved deep into how AI is not just evolving but revolutionizing the way we perceive and interact with technology.

Huang shared insights on NVIDIA's significant achievements in AI over the past decade, emphasizing the recent breakthroughs in generative AI. This transformative journey, as he described, has seen AI evolve to understand and interpret data in myriad forms, from languages to environmental physics. He especially highlighted the versatility of AI in translating and comprehending different 'languages', whether they are actual spoken languages or other forms of data representation.

A pivotal element of the discussion was Singapore's NAIS 2.0. Huang praised the strategic recognition of AI as vitally important and not just an opportunity. He highlighted Singapore's unique mix of languages and cultures, emphasizing the need for a specialized AI model like the SEA-LION (Southeast Asian Languages In One Network) Large Language Model (LLM). This endeavor is designed to craft a nuanced language model that caters to Southeast Asia's varied linguistic environment.

The purpose of SEA-LION is to address the biases present in existing LLMs, which often reflect the cultural values, political beliefs, and social attitudes of predominantly Western, Educated, Industrialized, Rich, Democratic (WEIRD) societies. These biases arise from the WEIRD-centric nature of internet data, which existing LLMs are predominantly trained on. In contrast, SEA-LION seeks to enhance representation of non-WEIRD populations and underrepresented languages in the Southeast Asian region, as reflected in its diverse training data distribution.

The meeting with Huang was more than just a discussion on technological advancements; it was a strategic session on aligning and propelling Singapore's AI ambitions with global technological trends. Huang's visit, set against the backdrop of Singapore's AI strategy announcement, underlines the integral role of AI in shaping not just the technology sector but the very fabric of societies and economies globally.

As the roundtable progressed, Jensen with his characteristic blend of expertise and enthusiasm, addressed a series of questions from the media, shedding light on various aspects of NVIDIA's role in the AI landscape and its collaboration with Singapore.

Singapore: A Hub for AI Excellence:

Huang described Singapore as an ideal launchpad for AI innovations. With NVIDIA's aim to be a leader in AI, the country serves as a center of excellence. NVIDIA's presence in Singapore for the past seven years has been marked by collaborations with researchers, venture capitalists, and a thriving startup ecosystem.

The Significance of the SEA-LION Project:

Delving into the SEA-LION project, Huang expressed his excitement about this unique collaboration. He stressed the importance of creating an AI model that truly resonates with the linguistic and cultural diversity of Southeast Asia, breaking away from the dominance of Western-centric AI models.

Supercomputing and Infrastructure:

In the session, Jensen suggested the potential construction of a new supercomputer in Singapore. However, when pressed by the media for confirmation, he clarified that every country needs its own supercomputer to develop foundational AI models. He emphasized that if Singapore doesn't take this initiative, it's unlikely that anyone else will. Jensen likened the necessity of a supercomputer for AI development to the need for specialized tools in other industries, such as telescopes in astronomy or unique lab equipment in scientific research, underscoring that high-speed, large-scale computing is essential for AI advancements.


When questioned about his concerns regarding competition, Jensen confidently responded that global competition is intensifying, and NVIDIA must continually innovate to maintain its lead. He acknowledged the robust competition from established players like Intel and AMD, as well as emerging companies from the East, such as Huawei, and 20-30 startups with a sharp focus on this market segment. However, he highlighted NVIDIA's unique position as a comprehensive provider, offering an array of products from chips and networking to software solutions. This broad spectrum of offerings, particularly in cloud computing – a domain where many startups rely on NVIDIA's technology – gives them a significant edge. Researchers are accustomed to and skilled in using NVIDIA's technology, making the switch to another provider a potentially daunting barrier.

 AI in less profitable sectors:

 I raised a question about the application of AI in sectors that are traditionally less profitable, like education and public health. Jensen quickly pointed to open-source models like Lama and Lama 2 as examples. However, he also noted that the choice of a Large Language Model (LLM) depends on whether the needs are general or specific. It's evident from his response that the development of large language models is on the rise, and researchers will increasingly need to explore innovative ways to optimize these models and maximize their potential.

Which leads to some of the key takeaways I got from the session: 

  1. In discussing AI and Large Language Models (LLMs), Jensen often referred to the concept of 'language', but his reference extended beyond just spoken or written words. He was alluding to the 'language of life and data' – essentially, how we can translate complex knowledge into binary code (1s and 0s). An apt analogy is considering protons, neutrons, and electrons as elements of a language. These fundamental components can be interpreted and processed by LLMs, enabling deeper analysis and understanding. 

  2. For governments to fully leverage AI and machine learning, having their own dedicated systems is crucial. This allows for the incorporation of localized data, encompassing unique aspects such as native languages, cultural nuances, and lifestyle patterns. Such tailored data inputs enhance the likelihood of successful AI applications while also ensuring local control and relevance. Naturally this would also benefit greatly all compute providers of which NVIDIA is currently the best option, but competition is growing and more opportunities arise. 

  3. Singapore is in a very strong position with tremendous opportunities in this field. The NAIS 2.0 and upcoming investments are a good starting point and will pave the way for innovation and with time can lead to become competitive with the leaders.

  4. Jensen, at his core, is just like any of us – he eats, he sleeps and he works. But it's his strong work ethic that probably plays a significant role in his success. The event wrapped up nicely with a light-hearted question about which hawker center in Singapore he should try, sparking many suggestions. This is so on brand for Singapore. They are all good and we’re looking forward to welcoming him back in the future.


bottom of page, pub-6094549887784613, DIRECT, f08c47fec0942fa0