Despite Silicon Valley claiming it welcomes the competition, the stock market reaction showed that the Hangzhou-based company is seen as a threat to the edge the US still enjoys.

In much of the world, however, it is viewed as an opportunity to allow local AI players to narrow the gap on US giants such as OpenAI or Google’s Gemini.

The importance of this cannot be overstated.

For the past two years, it had looked like large language models (LLMs) would remain forever out of reach of emerging economies because of limited human and financial resources.

Just a tiny number of companies in developed economies, hand in glove with governments, could aspire to ever play in the top league of AI development.

Worse yet, because of the way AI turbocharges research and innovation, any country left behind in the AI race was also set to lose out in a wide range of industries.

A gloomy prospect for countries like ours.

Equally alarming was that AI, it seemed, would allow the frontrunners to pull further away in areas crucial for national security, such as military technology and influence operations.

Rather than being just another industry, AI is a development multiplier. From a defense perspective, some may call it a superweapon.

Indonesia, along with the rest of the developing world, looked condemned to be spectators as AI raised the stakes in the US-China tech rivalry. Even Europe and Japan struggled to keep up.

Now, we can hope for more, for two reasons.

First, DeepSeek has developed an algorithmic model that reportedly requires much less data and computing resources than those devised in Silicon Valley to produce similar results, notwithstanding some doubts over claims regarding the Chinese company’s investment.

More importantly, DeepSeek’s LLMs are open source, meaning anyone can study their source code and copy and adapt it for their own purposes, free of charge. In that sense, DeepSeek puts OpenAI to shame for failing to live up to its name.

While the first of those two reasons means a hypothetical Jakarta-based start-up could build an AI platform without the gigantic investment Silicon Valley powerhouses have made to buy top-line processors from chip giant Nvidia, the second means it could use a blueprint to get going.

Together, these two factors give us the chance to catch up, or at least to narrow the gap, in the AI race.

But that is about it. Nothing about that is automatic.

It may well be that we remain on the sidelines, a paying user rather than a gaining creator of this frontier technology, and likely without any access to its cutting edge.

DeepSeek gives up a leg up, but we have to take the next step by developing the human resources and infrastructure needed.

This will require active support from the government. Both Washington and Beijing have supported, no, pushed, their respective AI industry through massive subsidies, favorable regulations and investment in research and development.

Jakarta, too, must create the academic, financial and legal foundation to attract venture capital to the AI field. Eventually, we will also need to work on the hardware side to develop chipmaking skills within our borders.

Government statements will not speak that industry into existence.

The significance of semiconductors, AI and machine learning for our overall national development, and for achieving our 2045 vision, is well understood in the tech world, but how about the political world?