Chinese AI Sends Ripples Through the Golbal Tech World

The rapid rise of DeepSeek, a Chinese artificial intelligence (AI) company founded in 2023, is sending ripples through the global tech industry, with significant implications for emerging economies like Indonesia. Analysts and local tech leaders suggest that DeepSeek’s cost-effective, open-source large language model (LLM) could lower barriers to AI adoption in the archipelago, potentially transforming how local startups and developers approach innovation. However, challenges such as inadequate infrastructure and a shortage of skilled talent remain substantial hurdles to fully capitalising on this technological wave.

DeepSeek’s recent releases, including the powerful V3 model in December 2024 and the reasoning-focused R1 model on 20 January 2025, have positioned the company as a serious competitor to industry giants like OpenAI and Anthropic. With performance reportedly rivalling OpenAI’s GPT-4o and Claude 3.5, but at a fraction of the training cost, DeepSeek is democratising access to high-quality AI tools. This development is particularly relevant for countries like Indonesia, where budget constraints often limit technological advancement.

A New Era for Indonesian AI Development

The potential impact of DeepSeek’s technology on Indonesia’s tech ecosystem is profound. Irzan Raditya, CEO of Jakarta-based conversational AI firm Kata.ai, highlights that the lower computing costs associated with DeepSeek’s models could empower local developers to experiment and build prototypes tailored to Indonesia’s unique needs. “This is a chance to create solutions that resonate with our market,” he stated.

Sofian Hadiwijaya, chief technology officer at AI virtual chat developer AvataraLabs, echoes this optimism. He notes that the cost of using DeepSeek’s models is remarkably low, with calculations suggesting a per-chat cost of just Rp 30 (less than 1 US cent) for users. “We’re moving closer to a reality where AI computation costs could be negligible, as Nvidia CEO Jensen Huang once envisioned,” Sofian remarked to The Jakarta Post on 6 February 2025. This affordability is a stark contrast to previous pricing structures in the industry, with DeepSeek’s offerings reportedly costing a quarter of OpenAI’s rates, even as competitors like OpenAI and Google slash their own prices in response—OpenAI recently launched its o3-mini model at half the price of its o3 version.

For a country with a burgeoning startup scene, such cost reductions could be transformative. Roshan Raj, Southeast Asia partner at Redseer Strategy Consultants, describes DeepSeek’s emergence as a “wake-up call” for emerging economies. “It shows that technological breakthroughs are possible even with capital and infrastructure constraints, provided you have high-quality tech teams,” he told The Jakarta Post on 6 February 2025. Raj believes this levels the playing field, giving countries like Indonesia a credible shot at competing in the global AI landscape.

Infrastructure: The Missing Piece of the Puzzle

Despite the excitement surrounding DeepSeek’s offerings, significant obstacles remain for Indonesia to harness this potential fully. A critical issue is the lack of high-performance graphics processing unit (GPU) infrastructure, essential for training and refining AI models. Sofian of AvataraLabs points out that even DeepSeek, with its cost-efficient approach, relied on 2,048 Nvidia H800 chips to train its V3 model. The company’s founder, Liang Wenfeng, revealed in a 2023 interview with the South China Morning Post that DeepSeek had amassed over 10,000 Nvidia GPUs, making it one of the largest holders of computing power among Chinese AI startups.

In contrast, Indonesia’s infrastructure lags far behind. Sofian referenced the recent inauguration of an AI Center at Brawijaya University in Malang, East Java, by the communications and digital minister. However, details about the facility’s GPU capacity remain unclear, raising questions about its ability to support cutting-edge research. “Without significant investment in high-performance computing, we cannot close the gap with AI powerhouses like China and the United States,” Sofian cautioned.

Irzan of Kata.ai agrees, stressing that AI research demands not just sophisticated computing power but also large-scale, affordable, and clean-energy-driven solutions. “This is where government and industry consortia must step in to provide resources,” he urged. He suggests that Indonesia could look to models like Saudi Arabia’s investment in 3,000 Nvidia H100 chips—each costing around US$40,000—at the King Abdullah University of Science and Technology as a blueprint for building capacity.

Talent and Investment: Building a Sustainable Ecosystem

Beyond hardware, the shortage of local talent and funding for AI research poses another challenge. While DeepSeek’s open-source LLM offers a starting point, translating this into meaningful innovation requires skilled professionals capable of adapting and building upon these tools. Roshan Raj advocates for strategic partnerships with countries more advanced in AI development, alongside investments in local tech talent. “Enticing talented Indonesians abroad to return and contribute to local AI development could be a game-changer,” he noted.

Moreover, positioning Indonesia as an attractive hub for AI research and investment is critical. This involves not only financial incentives but also creating an ecosystem where startups can thrive. Irzan advises local firms to focus on real-world applications rather than getting caught up in the rapid evolution of specific models. “End users don’t care which LLM you use, just as they don’t ask about your cloud service. They only care if the solution meets their needs,” he explained. His advice is to prioritise the interface and application layer, ensuring AI tools are user-friendly and relevant to local contexts.

Global Competition and Local Opportunities

DeepSeek’s aggressive expansion has not gone unnoticed by global tech giants. OpenAI’s launch of the cost-efficient o3-mini model on 6 February 2025 is a direct response to the competitive pressure from DeepSeek. Similarly, Google has halved the cost of access to its Gemini API compared to OpenAI’s pricing in 2023. This price war benefits emerging markets like Indonesia, where affordability is key to adoption, but it also underscores the fast-paced nature of the AI industry, where staying relevant requires constant adaptation.

For Indonesia, the rise of DeepSeek represents both an opportunity and a challenge. On one hand, access to affordable, high-quality AI tools could spur innovation and economic growth, particularly in sectors like education, healthcare, and e-commerce, where AI applications can address local needs. On the other hand, without strategic investments in infrastructure, talent, and policy support, the country risks falling further behind in the global AI race.

A Call to Action and Opportunity to Deliver

The democratisation of AI through companies like DeepSeek offers a rare window for Indonesia to leapfrog technological barriers. However, turning this potential into reality demands a coordinated effort from government, industry, and academia. Building high-performance computing infrastructure, fostering local talent, and forging international partnerships are essential steps to ensure that Indonesia does not merely consume AI technology but actively contributes to its development.

As Roshan Raj aptly put it, DeepSeek’s success challenges the notion that only resource-rich nations can lead in AI. For Indonesia, the message is clear: with the right investments and policies, the country can carve out a space in the global tech landscape. Whether this translates into tangible progress remains to be seen, but the arrival of cost-effective tools like DeepSeek’s models marks a pivotal moment for Indonesia’s digital future.