[Reporter’s Notebook] One Year After the DeepSeek Shock: South Korea Scrambles for Originality
One year ago, the Chinese artificial intelligence (AI) startup DeepSeek shocked the world by releasing an open-source inference model. DeepSeek, which achieved high-performance AI with low cost and low efficiency without advanced hardware infrastructure, has become synonymous with cost-effective AI models.
Korea sprang into action. The government, recognizing the need to secure domestic technological capabilities, launched its own AI foundation model project. With the determination to establish a "top three in AI" system within a short period, it has been leading efforts such as providing support for graphics processing units (GPUs).
However, from the outset, whether the project would be "from scratch" (developed independently from the beginning) became a subject of controversy. As the government emphasized "independence," a negative perception of open-source utilization spread throughout the industry. The lack of clear guidelines led to confusion over what could truly be considered an independent model. Naver Cloud failed the first-round evaluation for not meeting the criteria, and major companies subsequently announced their intention not to participate in the renewed bid.
In reality, DeepSeek has been actively leveraging open source to maximize efficiency. During the model development process, DeepSeek accelerated its progress by adopting modules from companies like Google and OpenAI, later unveiling ultra-fast, long-form training and inference technology that surpassed existing mechanisms, and introducing innovations such as sparse attention technology to reduce inference costs.
This is why some in the industry point out that making full use of Korea's outstanding refinement capabilities would be advantageous in terms of both time and cost. There is also skepticism toward developing fully independent AI models. The concern is that focusing solely on domestic use could hinder the ability to secure global competitiveness. At the same time, it is suggested that there is a greater need to focus on vertical (specialized) models rather than general-purpose ones.
Hot Picks Today
"Two Spend 500,000 Won in a Flash, Yet Wait Two Hours?" ... 'Dreamland' Shunned by Kids and Foreign Tourists
- The Era of Getting 200 Million Won After 30 Years Is Over... A Japanese Bank's "Perpetual Department Head" Who Just Clocked In Cries Out
- Spain Hit by Its Worst Tragedy... Lee Kangin, Moreno and Real Madrid Join Soccer World in Mourning
- They Said, “Make It Your Home for Life” - Then Suddenly Kicked 200 Passengers Off... What Happened to the Around-the-World Cruise?
- "130 Rice Bowls Sold Out in a Day"... 'Unfiltered Face' Adds 8 Million Won to Night Market Sales: China's Bungeunwoo
There are certainly fields where data sovereignty is required. However, we now live in an era where companies are actively adopting ChatGPT and individuals freely use and share their experiences with ChatGPT or Gemini. The two AI models from Google and OpenAI, which are engaged in fierce development and update competition, have already permeated everyday use. The core question for AI models is how widely they can be used. While independence is important, it is more critical to harness open-source resources effectively to improve performance and ensure controllability.
© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.