That a bootstrapped Chinese startup, DeepSeek, burst onto the international tech scene and is giving industry giants like Google and OpenAI a run for their money, is making AI creators across the development spectrum rethink their budgets, and ranking of supremacy of AI models vis-a-vis functionality.
DeepSeek, a Chinese AI company based in Hangzhou, was founded towards the end of 2023 by Liang Wenfeng, a 40-year-old hedge fund manager. Liang utilized gains accrued from High-Flyer Quantitative Investment Management, his hedge fund, to finance his AI agenda, making quiet strides, and now his work is doing the ‘talking’.
High-Flyer created computer programs to trade stocks and later used machine learning to improve those trading strategies. The company’s strong research and development system receives a large annual investment towards that direction, which translates to regular tests and improved strategies, giving the technology a unique competitive edge.
DeepSeek released its free chatbot app based on DeepSeek – R1 model for Apple IOS and Android on 10th January 2025, catapulting its number of downloads to the number 1 spot in the United States in just over 2 weeks. The quest for AI dominance is fast taking shape, and has seen a number of market players take financial hits.
By 27th January, global tech investor-stokes took a dip, with Nvidia, a respected market leader, taking the biggest punch on Wall Street – a record $593 billion in one day.
Using less data, at a fraction of the cost of other service providers, got tech enthusiasts jumping ship to the new kid on the block. And the tech giants’ loss had a ripple effect on their suppliers, subsidiaries, partners, and industry stakeholders alike.
Take for instance Advantest, a chip-testing maker and supplier to Nvidia, which took a 10% dive. Or chip-making equipment developer Tokyo Electronic that went down 5%. Or the data centers across major global capitals that also took a hit.
Have the AI tech giants’ equities and stock market values been overly hyped and inflated? Companies like OpenAI, Google, and DeepMind have created very advanced AI models that need huge amounts of investment and powerful computer systems to run.
The belief has been that making AI better always requires spending more money. DeepSeek has shown that is not true. What makes DeepSeek stand out is that it works just as well as other models, but costs much less to operate.
With OpenAI for example, $100 is required for every million tokens, while for DeepSeek, only $4 is required to operate, a massive $96 drop in cost. Plus, unlike OpenAI and Google’s closed-source models, DeepSeek is open-source and has a permissive license, so developers can freely use and build on it without limits.
The AI industry is undergoing a major shake-up, taking a turn from throwing more resources at the current development approach – massive computing power, energy and money. The focus is shifting to smarter leaner approaches that make AI more efficient and widely usable.
Companies that embrace this shift – those that stay nimble – will lead the next era of AI advancements. Companies that do not evolve will ultimately be left behind and struggle to remain relevant.
DeepSeek trains its R1 AI models differently than Open AI. Their method takes less time, uses fewer specialized chips, and costs way less money. The company’s long-term goal is to build Artificial General Intelligence (AGI) – AI that can think and learn with humans. Their improvements in how AI reasons and solves problems are a huge step forward for the field.
According to a Reuters report of 25th January 2025, Liang Wenfeng, DeepSeek’s founder, was invited to share his opinion on Chinese government policy, a move that highlights Beijing’s recognition of DeepSeek’s role in potentially upending the global AI order, in China’s favour.
“China’s AI can’t be in the position of following forever. We often say that there is a gap of one or two years between China’s AI and the United States, but the real gap is the difference originality and limitation.” – Liang Wenfeng
DeepSeek is democratizing access to AI technology, making it more accessible to developers worldwide. The company’s success is a wake-up call for the US to re-evaluate its AI strategies and policies. The company’s open-source approach is a more collaborative and inclusive way of developing AI.
DeepSeek’s rise also has geopolitical implications and is a sign of China’s growing technological prowess.
By Alfred Galandi
