🔗 Share this article

Internationally, nations are channeling hundreds of billions into the concept of “sovereign AI” – creating domestic AI technologies. Starting with the city-state of Singapore to Malaysia and the Swiss Confederation, nations are competing to develop AI that understands native tongues and local customs.

The Worldwide AI Arms Race

This trend is a component of a broader international competition led by major corporations from the America and China. While companies like OpenAI and a social media giant allocate massive funds, middle powers are likewise making sovereign investments in the AI field.

Yet amid such vast sums in play, can smaller nations attain significant gains? According to an expert from a prominent research institute, “Unless you’re a wealthy government or a large company, it’s a significant hardship to build an LLM from the ground up.”

Security Concerns

Many nations are hesitant to rely on overseas AI technologies. Throughout the Indian subcontinent, for example, American-made AI systems have occasionally fallen short. One example saw an AI assistant employed to instruct pupils in a isolated village – it interacted in the English language with a thick American accent that was nearly-incomprehensible for regional listeners.

Furthermore there’s the defence dimension. In India’s security agencies, relying on particular international AI tools is viewed not permissible. According to a founder commented, “It could have some random training dataset that may state that, for example, Ladakh is outside of India … Utilizing that particular model in a security environment is a big no-no.”

He continued, I’ve consulted people who are in defence. They wish to use AI, but, forget about certain models, they are reluctant to rely on Western systems because information might go abroad, and that is totally inappropriate with them.”

Domestic Projects

Consequently, some nations are backing national projects. A particular such effort is underway in India, where an organization is working to create a domestic LLM with state backing. This project has committed roughly a substantial sum to machine learning progress.

The founder envisions a system that is more compact than top-tier tools from American and Asian firms. He notes that India will have to compensate for the financial disparity with expertise. Located in India, we don’t have the luxury of investing massive funds into it,” he says. “How do we vie versus for example the hundreds of billions that the United States is pumping in? I think that is where the fundamental knowledge and the strategic thinking comes in.”

Native Priority

Across Singapore, a government initiative is backing AI systems trained in south-east Asia’s local dialects. Such languages – including Malay, Thai, Lao, Indonesian, Khmer and others – are commonly underrepresented in US and Chinese LLMs.

It is my desire that the people who are developing these independent AI tools were conscious of just how far and just how fast the leading edge is moving.

A senior director participating in the program notes that these tools are intended to supplement larger AI, as opposed to displacing them. Systems such as ChatGPT and Gemini, he says, commonly have difficulty with native tongues and culture – speaking in stilted the Khmer language, for example, or suggesting meat-containing meals to Malay individuals.

Building local-language LLMs permits state agencies to include cultural nuance – and at least be “smart consumers” of a sophisticated tool developed elsewhere.

He adds, I am prudent with the concept independent. I think what we’re aiming to convey is we aim to be more accurately reflected and we want to understand the abilities” of AI platforms.

Multinational Partnership

Regarding countries trying to establish a position in an growing worldwide landscape, there’s a different approach: team up. Researchers affiliated with a well-known institution put forward a public AI company allocated across a alliance of developing nations.

They term the project “an AI equivalent of Airbus”, in reference to the European effective play to develop a rival to Boeing in the mid-20th century. The plan would entail the establishment of a government-supported AI organization that would merge the assets of various nations’ AI projects – such as the United Kingdom, Spain, the Canadian government, Germany, the nation of Japan, the Republic of Singapore, South Korea, France, the Swiss Confederation and Sweden – to develop a competitive rival to the Western and Eastern giants.

The main proponent of a study setting out the concept states that the concept has attracted the consideration of AI ministers of at least a few countries to date, in addition to several sovereign AI organizations. Although it is presently focused on “developing countries”, less wealthy nations – Mongolia and Rwanda included – have also expressed interest.

He explains, Currently, I think it’s an accepted truth there’s reduced confidence in the assurances of the present American government. People are asking like, should we trust these technologies? In case they opt to