Why DeepSeek’s success is worrying Silicon Valley

Chinese-made new artificial intelligence (AI) model DeepSeek has taken the tech world by storm. It was released in January 2024 and the app quickly soared to the top of Apple’s App Store downloads. The phenomenon struck headlines for its cost-effectiveness and impressive capabilities.

DeepSeek is apart from competitors due to its ability to perform as well as established models like OpenAI’s GPT-4 and at a fraction of the cost. The cost efficiency is largely due to the use of less expensive chips and this has caught the attention of AI experts as well as investors. The company behind the app claims that it had trained the AI with only $6 million.

The breakthrough is also a challenge to the efforts of Washington to curb the rise of China in the tech world. Beijing has doubled down on its ambition to lead in AI. DeepSeek and more such Chinese startups is at the forefront of this technological push.

DeepSeek is not without controversies despite the excitement. It has been trained to avoid politically sensitive topics. The app responded with a standard disclaimer and reflected China’s strict censorship laws when asked about the Tiananmen Square massacre.

DeepSeek founder Liang Wenfeng is a relatively unknown figure who has quickly gained prominence.

The development is a double-edged sword for inestors. Companies like Nvidia witnessed a sharp drop in their stock prices equipped with some estimates placing the loss at nearly $600 billion in a single day.

It is clear that we are witnessing the rise of a new era in AI as DeepSeek continues to disrupt the status quo.

Q&A

Q1: How does DeepSeek compare to other AI models like OpenAI’s GPT-4?

A1: DeepSeek has been designed to be as powerful as OpenAI’s GPT-4, but at a much lower cost.

Q2: What makes DeepSeek’s rise so significant for the AI industry?

A2: DeepSeek’s rapid rise is noteworthy because it challenges the conventional belief that AI models must be built with expensive, top-tier chips.

Q3: How has DeepSeek impacted the stock market?

A3: DeepSeek’s entry into the AI market and its demonstrated cost-efficiency have led to significant financial repercussions for major tech companies like Nvidia. Nvidia’s stock dropped nearly 17% after DeepSeek’s rise, as investors questioned whether companies relying on high-end chips for AI innovation could continue to justify their massive valuations without offering significant returns.

Q4: What are the concerns surrounding DeepSeek’s use of censorship?

A4: One of the concerns about DeepSeek is that it is trained to avoid politically sensitive topics, reflecting China’s strict censorship policies. For example, when asked about the Tiananmen Square incident, DeepSeek refused to answer, highlighting the limitations placed on AI in China.

Q5: How did DeepSeek manage to develop its AI model at a fraction of the cost of competitors?

A5: DeepSeek was able to develop its model cheaply by using a combination of less advanced chips and gathering a stockpile of Nvidia A100 chips (which had been banned for export to China) to reduce the cost of AI training. The company claims it only spent around $6 million to train its model, compared to the hundreds of millions that companies like OpenAI have spent on their own models.

Q6: Who is Liang Wenfeng, the founder of DeepSeek, and what is his background?

A6: Liang Wenfeng is the founder of DeepSeek and has a background in both finance and AI.

Q7: What does DeepSeek’s rise mean for the future of AI development?

A7: DeepSeek’s success challenges the idea that AI development is limited to companies with vast financial resources and access to the most advanced chips. It suggests that the future of AI may see more competition from innovative startups capable of developing powerful models on a smaller budget. This could democratize AI technology, making it more accessible and prompting major companies to reconsider their strategies in both hardware and software development.