DeepSeek, China’s Hangzhou-based AI startup cum research lab, launches its disruptive open-source AI Model DeepSeek R1 generating global tremors across the tech markets and app stores. Within a few days after DeepSeek R1’s launch, it already tops the chart on the Apple App Store surpassing OpenAI’s ChatGPT in the US, UK, and China regions. DeepSeek R1, although a cost-effective model, outperforms OpenAI’s latest reasoning model OpenAI-o1 in several Maths and reasoning metrics. 

As per the claims of the company, it has been developed for just $ 5.6 Million whereas GPT incurred a whopping $5 Billion just last year. It stands out even for offering unlimited free usage making it a go-to AI destination for enthusiasts. Apart from all these, the model is also distinct in its approach as it delivers the reasoning first and the response to the user prompt later. Also, that’s why it remains in stark comparison with Open AI’s latest model OpenAI-o1 which also works in somewhat the same manner focusing more on the quality of the response. 

DeepSeek: The AI Race 

As mentioned about the investment and response on App stores, the AI model has sent tremors in Silicon Valley, specifically to the AI giants like Open AI, Meta, and Nvidia among others. Moving beyond the response received from the market, in its use regarding GPUs, the company has claimed to use only a fraction of chips as compared to Meta’s Llama, Anthropic, or even OpenAI to train its AI model. This brings us to the most interesting part where the market remains seriously concerned about whether AI companies really need billions of funding and multifold billions of evaluations. 

Read about Anthropic's latest addition at: Anthropic’s Citations: This New Feature Will Make Sure You Get Relevant and Accurate Responses from AI Every Time, Check Out How!

DeepSeek: Pushing Sell-offs 

The concerned question is quite visible in the response that the markets have received in the last few days with the investors probably searching for answers and taking a back seat. For instance, Nvidia’s stock price dipped more than 17% on Monday morning, chipmaker ASML went down nearly 6% while the Nasdaq dropped more than 3%.  

This marks an inflection point in the global AI race where despite stringent restrictions posed by the US on the availability of Chips and associated technology to AI companies in China, a Chinese startup has managed to outperform Silicon Valley’s biggest AI giants raising plenty of questions on the strategies and vision of the AI giants like OpenAI, Meta and Nvidia among others. This also raises serious questions regarding the effectiveness of the strategy followed by the US aimed at restraining China. 

DeepSeek: Unresponsive Questions 

In a bit of surprise yet conforming to the unsaid norms, the model denies replying to questions that can potentially involve drawing comments upon Chinese aggression and anti-social practices that have previously happened or are ongoing anywhere around the world. It includes questions on Tiananmen Square and the Arunachal Dispute among others. 

Read about ChatGPT's Operator, Computer Using Agent, at: Test Scores of ChatGPT’s All-new Computer-Using Agent “Operator” Might Blow Your Minds

It highlights a shift in the AI landscape, where innovation can emerge rapidly and at lower costs than previously anticipated. This brings a milestone in the AI space where pure innovation and some good strategies, combined, have delivered an all-surpassing AI model.