Samsung is putting all its might into developing its own artificial intelligence technology

South Korean tech giant Samsung is working tirelessly to advance its own technology, to avoid falling into the hands of its competitors in the future.

A recent report suggests that Samsung Electronics has embarked on a full-throttle initiative to develop its proprietary Large Language Model (LLM). This strategic move was triggered by an unfortunate leak of internal data. Samsung Research spearheads the project, with considerable investments allocated towards manpower and resources. The tech giant has allocated all its available GPU resources for the development of the LLM, a crucial factor in training these models. The goal is to roll out the first version of the LLM within the next two months.

The primary application of Samsung's impending LLM solution is expected to revolve around document summarization, software development, and language translation. While there is currently no decision regarding whether the AI solution will be made accessible to the public, its internal utilization is regarded as essential. The recent data leak, which arose due to the use of third-party generative AI solutions like ChatGPT, led Samsung to prohibit external solutions, prompting the development of its own AI technology.

Samsung is putting all its might into developing its own artificial intelligence technology

Samsung is optimistic that the implementation of LLM-based AI solutions can significantly expedite processes in software development and semiconductor design. Although collaborations with prominent entities such as Google, Microsoft, Naver, and OpenAI were considered, the top brass, including Lee Jae-yong, ultimately decided to develop an in-house AI solution.

In South Korea, several firms including Kakao, LG, Naver, and SKT, are concurrently developing their LLM-based generative AI solutions like KoGPT, ExaOne, HyperCloverX, and A-Dot. These solutions are designed to emulate human-like drawing and writing skills. However, these initiatives face significant challenges related to data accuracy and privacy protection, issues that must be addressed during the development phase.