Singapore Announces S$70 Million AI Initiative to Develop Southeast Asia’s First LLM
Singapore plans to develop a large language AI model and train it, to understand and generate languages from Southeast Asia.
For the same, Singapore announced a new S$70 million initiative – the National Multimodal LLM Programme (NMLP) – that could serve as the foundation for various text-to-speech or text-to-image generative systems, and develop the country’s research and engineering capabilities for multi-modal large language models.
According to the announcement, the initiative was launched by the Infocomm Media Development Authority (IMDA) in partnership with AI Singapore (AISG) and the Agency for Science, Technology and Research (A*STAR).
Funded by the National Research Foundation, the program aligns with Singapore’s National AI Strategy 2.0, unveiled by Deputy Prime Minister Lawrence Wong at the Singapore Conference on AI yesterday.
“This national effort underscores Singapore’s commitment to become a global AI hub,” said Dr Ong Chen Hui, assistant chief executive of biztech group at IMDA. “Language is an essential enabler for collaboration. By investing in talent and investing in large language AI models for regional languages, we want to foster industry collaboration across borders and drive the next wave of AI innovation in Southeast Asia.”
Singapore’s Plans to Upend Western Dominance in LLMs
Addressing the current void in the scene, where most LLMs hail from Western countries, the National Multimodel LLM Programme aims to contribute to Singapore’s capabilities in AI research and innovation.
It seeks to build skilled AI talent by providing funding and access to high-end computing for local researchers and engineers while fostering the AI industry to develop LLM-enabled solutions for enhanced productivity and new business opportunities.
The LLM project will also play a role in establishing a trusted environment for AI usage. By providing a deeper understanding of LLMs and furthering research in AI governance, the initiative aims to ensure responsible and ethical AI applications. It builds upon the success of AI Singapore’s Southeast Asian Languages in One Network (Sea-Lion) model, an open-source LLM designed to be smaller, faster and more flexible than its mainstream counterparts.
The Sea-Lion model, initially trained on 11 languages used in the region, including English, Chinese, Indonesian, Malay, Thai, and Vietnamese, served as a cost-effective and efficient alternative to existing options. The new national initiative seeks to expand Sea-Lion to encompass 30 billion to 50 billion parameters, transforming it into a multimodal speech-text model.
In comparison, well-known LLMs such as GPT-4 from OpenAI, boast 1,700 billion parameters and Meta’s Llama-2 consists of 70 billion parameters. However, the focus on cultural representation and linguistic nuances sets the Sea-Lion model apart, enabling users to understand the context and values of the diverse cultures and languages of Southeast Asia, particularly in the multilingual landscape of Singapore.
In line with the Trust Project guidelines, please note that the information provided on this page is not intended to be and should not be interpreted as legal, tax, investment, financial, or any other form of advice. It is important to only invest what you can afford to lose and to seek independent financial advice if you have any doubts. For further information, we suggest referring to the terms and conditions as well as the help and support pages provided by the issuer or advertiser. MetaversePost is committed to accurate, unbiased reporting, but market conditions are subject to change without notice.