Why Chinese AI company DeepSeek has Silicon Valley spooked
What's the story
Chinese artificial intelligence (AI) company DeepSeek has created quite a stir in the US tech industry. DeepSeek's AI assistant recently dethroned OpenAI's ChatGPT to become the highest-rated free app on Apple's App Store in the US. The development comes as a major milestone for the Hangzhou-based start-up, which was founded in May 2023 by billionaire Liang Wenfeng.
Swift ascent
Rapid rise in popularity
DeepSeek officially arrived in the US on January 10, and in a matter of just 17 days, it has gone viral. Its platform is based on the DeepSeek-V3 model, which its developers claim "tops the leaderboard among open-source models and rivals the most advanced closed-source models globally." This rapid ascent makes DeepSeek a strong competitor to the likes of OpenAI, Google, and Microsoft.
Market influence
Impact on global tech stocks
The meteoric rise of DeepSeek has also taken a toll on global tech stocks. The Nasdaq 100 futures declined by as much as 1.9%, while contracts on S&P 500 fell by up to 1%. This market response highlights the looming threat of emerging AI platforms such as DeepSeek to the dominance of established tech giants in the industry.
Competitive edge
DeepSeek's unique features
DeepSeek has already made a name for itself by matching or even outperforming leading US models. The buzz around DeepSeek is all about its cost-effectiveness and capability to run on reduced-capability chips. These features challenge the dominance of US companies such as NVIDIA in the AI space and question the high valuations of these firms.
Training efficiency
Training process and cost
Despite US government's export controls to prevent advanced chips from being used to train Chinese firms' AI models, DeepSeek has trained its model using NVIDIA's H800 chips for under $6 million. However, this claim remains hotly debated in the tech community.
Strategic shift
DeepSeek's impact on AI development strategies
The success of DeepSeek has ignited a debate on whether we are approaching AI development the right way. Nirgunan Tiruchelvam, chief of consumer and internet at Singapore-based Aletheia Capital, told Bloomberg that DeepSeek's product, "Is deeply problematic for the thesis that the significant capital expenditure and operating expenses that Silicon Valley has incurred is the most appropriate way to approach the AI trend."
Market disruption
Competitive edge and market impact
DeepSeek's AI model is viewed as a competitor to those of OpenAI and Meta Platforms. The company's assertion that it is much cheaper to train and develop has sent ripples through share prices across Asia's supply chain. This counters the idea that China's AI technology is behind its US counterparts, even with trade restrictions from Washington.
User experience
DeepSeek's transparency and cost-effectiveness
DeepSeek's assistant shows its work and reasoning as it answers a user's written query/prompt. This transparency has been appreciated by users. Notably, DeepSeek is very cost-effective, with its API priced at just $0.55 per million input tokens and $2.19 per million output tokens, as opposed to OpenAI's API which typically costs $20 per million tokens.