The writer is a founding partner of Hong Kong-based VC firm IN. Capital
DeepSeek has forever altered the trajectory of global rivalry in tech. In China, founder Liang Wenfeng has become a local champion. For a country where overseas degrees — especially those in the US — are still perceived as more prestigious than their domestic equivalent, students and parents have been stunned to discover that his artificial intelligence start-up’s research team were all educated domestically.
Beijing is more confident than ever in its pursuit of the technology. DeepSeek’s success undermines the barriers that have been created in the US-China tech war.
The Hangzhou-based company’s decision to release a low-cost, open-sourced AI model, alongside detailed disclosure of its training methods, means that everyone, from researchers in São Paulo to start-ups in Stockholm and doctors in Nairobi, can access state-of-the-art AI at little to no cost.
Within the Chinese start-up sector a chain reaction is taking place. New AI applications are being created. Competition is going to become more fierce. Risk appetite from early-stage venture investment is increasing. DeepSeek’s decision to pursue an open-source AI model is inspiring and putting pressure on others to do the same. The first to react was Alibaba’s Qwen team, which released Qwen2.5 as open source last month on the eve of Chinese new year.
This is a remarkable change. After the US start-up OpenAI released its generative AI model ChatGPT in late 2022, the global digital economy was edging towards control by a handful of tech giants. These players chase scale over efficiency — building ever-larger models that demand staggering compute, energy and capital while guarding their training methods as trade secrets.
Centralised, closed models create a dangerous feedback loop. The more data they amass, the more powerful they become, further marginalising anyone outside their gates. For consumers this means large fees, surrendered data and watching AI’s future unfold without meaningful participation.
The promise of DeepSeek’s reasoning R1 model lies in its adaptability. Being open sourced it can be tailored to local needs. It avoids redundant calculations using something called sparse neural network training, meaning that its efficiency reduces compute and energy needs by orders of magnitude.
This means that advanced AI can benefit the masses, not just the few. It proves that the technology is a commodity. Billions of dollars need not be wasted on competition between tech giants with closed models. AI’s value should not lie in proprietary models but in what we are all able to do with it.
As an investor, I am concerned that DeepSeek’s prominence might lead the US to opt for even tougher sanctions. In China, export restrictions of graphics processing units (GPUs) such as Nvidia’s powerful H100s have hindered start-up growth. Funding from foreign investors is limited due to compliance risk concerns. But the real danger lies in limiting access to global education and research collaboration, which stifles the global knowledge flow that is critical to sustaining progress. Talent can circumvent chip shortages, but erecting barriers to learning risks long-term stagnation.
Yet even additional US restrictions, conspiracy theories and smear campaigns targeting DeepSeek cannot change the reality that the Chinese start-up has put AI into the hands of humanity.
Against all the noise, let’s consider this as a moment in history. In 1440, Johannes Gutenberg brought Europe the printing press, an invention that broke the monopoly on knowledge previously held by elites. DeepSeek’s achievement joins this tradition of making information more accessible. Its low-cost reasoning model proves that AI can belong to everyone, not just those who are hoarding codes, chips and capital.
