I noticed that at the beginning of 2025, a quite interesting event occurred in the AI world. The Chinese company DeepSeek, which was founded only in May 2023 by Liang Wenfeng, released a free chatbot application based on their DeepSeek-R1 model. And interestingly, by the end of January, it became the most downloaded free app in the American App Store, surpassing ChatGPT. This happened less than two years after the company's creation.

The fact is, DeepSeek is based in Hangzhou and initially positions itself as a developer of open large language models. They made an unusual choice — they open-sourced their algorithms, models, and even training details for free use and modification. This allowed researchers and developers around the world to study and improve their technologies.

What happened next was even more interesting. The success of DeepSeek created a wave in the Chinese AI market. ByteDance, Tencent, Baidu, and Alibaba — all these giants were forced to lower their model prices to compete. But here’s the paradox: DeepSeek remains a profitable company, while its competitors are incurring losses from the price war.

This is possible because the company focuses exclusively on research and is not rushing to commercialize. Such an approach allows them to avoid the strictest provisions of Chinese AI regulations. They hire young talents from leading Chinese universities and specialists from other fields, expanding the capabilities of their models.

On a broader level, this is seen as a significant step in the development of high technology in Asia. DeepSeek helps retain talent in the region and reduces brain drain from China, India, and other countries. This is an important moment for balancing technological development worldwide.