SEPTEMBER 4 — Nvidia’s decision to acquire Hugging Face for approximately US$12.9 billion should command attention far beyond Silicon Valley.

For Malaysia and Asean, this is not merely another corporate acquisition in an overheated technology sector. 

It is another indication that artificial intelligence is rapidly becoming part of the strategic infrastructure of the twenty-first century. 

The same alert, indeed, warning, can be issued to Small and Medium Enterprises (SMEs) all over the country and the region of Southeast Asia at large. Why ?

Hugging Face has become one of the world’s most important platforms for developers to share, test, customise and deploy AI models. 

Nvidia says the Hugging Face platform is used by more than 18 million developers, researchers and creators, hosts more than three million models and 500,000 datasets, and serves more than 200,000 companies.

Nvidia is therefore buying much more than another software company. It is moving further along the AI value chain.

The company already occupies an extraordinarily powerful position in the semiconductor architecture underpinning advanced AI. 

By acquiring Hugging Face, Nvidia gains a much deeper presence in the community developing and distributing open AI models.

The significance is difficult to exaggerate.

There has been considerable discussion about whether AI represents another technology bubble. 

Some valuations may indeed be excessive. Some AI companies will certainly fail. Investments will be written off. 

Expectations will occasionally run ahead of actual commercial applications.

But confusing speculative excess surrounding AI with AI itself would be a serious analytical mistake.

The dot-com bubble did not mean the Internet was irrelevant.

Similarly, even if parts of today’s AI investment cycle eventually experience a substantial correction, the technological transformation will continue.

There is an even more important reason.

AI possesses something few previous technological waves enjoyed at such an early stage: the strategic endorsement of the world’s two most powerful states.

The United States and China increasingly regard AI not merely as another commercial technology but as an instrument of economic competitiveness, military capability, industrial productivity and national power. In the US, as it is China, the AI race is akin to the Sputnik Moment that triggered the beginning of an arms race between Soviet Union and the US in the 1950s. 

The world is still feeling the impact and legacy of the arms race.

To be sure, Washington wants to preserve American leadership in advanced semiconductors, computing infrastructure and frontier AI models. 

Beijing is simultaneously accelerating the development of domestic chips, models and AI ecosystems. Malaysia and Asean have to watch everything closely. 

A command center on AI’s latest development is needed not unlike the ticker of Bloomberg that guides the Stock Exchange in New York, for that matter, Morgan Stanley or Standards and Poor’s.

The result is effectively an AI amazing sprint—although one conducted through algorithms, computing power, electricity, data centres, semiconductors and human talent rather than merely conventional hard power.

China’s progress illustrates why the competition cannot be dismissed.

Companies such as DeepSeek even KIMI have demonstrated that Chinese developers can produce increasingly sophisticated models at lower costs. 

Interestingly, the Chinese developer of KIMI has a PhD from Carnegie Mellon University too. Jensen Huang himself has been encouraging the US not to ban the sale of any chips to China.

Huawei, meanwhile, has expanded rapidly in China’s AI semiconductor market as Beijing seeks greater technological self-reliance.

Indeed, Nvidia CEO Jensen Huang has acknowledged how dramatically the Chinese market has changed. 

Nvidia once commanded an overwhelming share of China’s advanced AI chip market, but Huawei and other domestic competitors have increasingly filled the space created by American technology restrictions.

This competition is precisely why Nvidia’s moves deserve attention.

When a company sitting near the centre of the world’s AI computing infrastructure spends almost US$13 billion acquiring one of the largest communities of open-model developers, Malaysia should not regard the development as distant Silicon Valley financial engineering.

It is a signal. An important one. 

The centre of competition is moving from merely producing better chips towards controlling—or at least influencing—the entire AI ecosystem: chips, models, developers, data, cloud infrastructure, applications and increasingly robotics.

It helps the SMEs of Malaysia and Asean when they can begin to use more and more of AWS to improve their efficiency. Even think tanks must use them to organize their research decks non stop. 

Malaysia and Asean cannot avoid this transformation.

Nor should they attempt to choose prematurely between American and Chinese technological ecosystems.

Asean’s advantage has traditionally rested in maintaining strategic flexibility. The same principle should apply to AI.

Malaysia should engage Nvidia, Microsoft, Google, Amazon and other American technology companies while simultaneously understanding developments involving Huawei, Alibaba, Tencent, DeepSeek and other Chinese AI players.

Strategic autonomy in the AI era will not come from technological isolation. It will come from technological literacy and diversification.

Malaysia already possesses several advantages.

Its semiconductor ecosystem is deeply embedded in global supply chains. Johor is becoming an important data-centre location. Malaysia has significant manufacturing capabilities and relatively sophisticated digital infrastructure.

But infrastructure alone is insufficient.

The country must develop AI engineers, researchers, programmers, data scientists and—equally importantly—workers capable of applying AI throughout the economy.

Universities must adapt. Government agencies must understand AI procurement. Malaysian companies, including SMEs, must learn how to integrate AI into logistics, manufacturing, healthcare, education, agriculture and services.

Asean should think similarly.

With more than 680 million people, the region should not resign itself to becoming merely a consumer market for American or Chinese AI.

Asean needs computing capacity, interoperable data systems, reliable electricity, regional digital infrastructure and significantly stronger AI education.

This connects AI directly with another major Asean strategic project: energy integration.

Data centres and advanced AI require enormous amounts of electricity. The Asean Power Grid therefore should increasingly be understood not merely as an energy project but as part of Southeast Asia’s future digital and AI architecture.

The countries capable of supplying reliable energy, computing infrastructure and skilled human capital will capture a greater portion of the emerging AI economy.

Nvidia’s acquisition of Hugging Face is therefore another reminder that the AI revolution is widening rather than narrowing.

Whether Nvidia ultimately wins every technological battle is beside the point.

The United States and China have already decided that AI is strategically indispensable.

That alone makes the technology extraordinarily difficult for the rest of the world to ignore.

Malaysia and Asean do not have to join the AI race between Washington and Beijing.

But they cannot afford to watch it from the sidelines either.

When Nvidia makes a move, Malaysia and Asean should pay attention.

* Phar Kim Beng, PhD is the Professor of Asean Studies at International Islamic University of Malaysia and Director of Institute of International and Asean Studies (IINTAS). 

** This is the personal opinion of the writers or publication and does not necessarily represent the views of Malay Mail.