DeepSeek R1 Disrupts AI Market

Al Jazeera

In January, DeepSeek, a Chinese artificial intelligence (AI) startup, released its latest model, DeepSeek R1, which rivals technology developed by OpenAI at a fraction of the cost. 

The market’s reaction rattled investors, as DeepSeek made headlines and caused many US tech stocks to sink. One of the sinking tech stocks is Nvidia, which has raised the question of whether American firms would continue to dominate the AI market. 

DeepSeek looks, feels, and acts similarly to ChatGPT by OpenAI. DeepSeek’s R1 is said to be just as powerful as their o1 model–released at the end of last year–which can perform complicated math and coding tasks. However, DeepSeek reports being able to train its model at a fraction of the cost–using only $6 million as compared to the “over $100m” referenced by OpenAI CEO Sam Altman when discussing GPT-4. 

Using less memory that rivals, DeepSeek attracts customers through its combination of lower cost and performance. In fact, it became the most-downloaded free app on Apple’s App Store when it was released in the US.

Nvidia’s shares tumbled by 17% on January 27th, recovering since then. While tech companies choose to spend aggressively to invest in AI, DeepSeek’s efficiency disapproves that idea. Meta, on the other hand, jumped 6.4%, likely due to its strong quarterly earnings report.