What is the reason why Deepseek created a shock?
Launched in a short time but DeepSeek has given himself a “terrible” achievement, storming with the technology world that US President Donald Trump also spoke out to “remind” American technology companies.
Starting from a Chinese technology company, Deepseek announced V3 and R1 is considered a “Sputnik moments” in the field of artificial intelligence, shocking the world similar to the first Soviet satellite launcher. October 4, 1957. Business Today also does not hesitate to express Deepseek being considered a turning point in the global AI battle, which can lose the position of the giants who have become famous in the US in the international technology school.
Deepseek low cost and quality
Back in 2022, the world market was vibrant with a lot of AI chatbot, starting to be chatgpt later to Gemini and many other technologies from Western countries. With the first explosion in 2025, the trend is gradually changing, China launched Deepseek with very low development costs, only a small part compared to the super -expensive AI models of American giants. The model also does not need to use the strongest chip clusters but still achieve superior effect.
Accordingly, the R1 model is trained on NVIDIA’s less advanced H800 chips, 100% open source, 96.4% cheaper than Openai O1 but still achieved equivalent efficiency. Production time is also in just 2 months and costs about 6 million USD. If compared to Openai O1, it takes 60 USD for one million output tokens, Deepseek R1 only needs $ 2.19.
Thus, the launch of Deepseek at a lower cost of 20 to 50 times higher than the GPT of Openai has proven that the advanced AI model can completely deploy with limited resources.
Open source code
Meta’s AI scientist, Yann Lecun, has a difference about Deepseek, which is open source models that are surpassing exclusive models. Lecun said that Deepseek benefits from open research.
According to CNBC, open source helps people access smart AI models easier, low cost but equivalent efficiency. Mr. Nguyen Hong Phuc, AI expert in Ho Chi Minh City, assesses the open source factor is still enough to help Deepseek earn profits, creating ecosystems that can collect money.
China and the United States confronts fair
Many technology experts said that China’s launch of DeepSeK showed that the potential for developing AI technology was growing stronger, gradually changing the direction of developing AI in the world. Although the polarization or the picture is still unclear, we can see China’s ambition in AI race.
Besides, DeepSeeK appears to inspire many technology companies to innovate and develop to launch new technologies.
Post Comment