In the coming year of 2025, generative artificial intelligence (GenAI) is expected to reach a critical turning point, where we will gain a clearer understanding of its true potential. As analyses of its various applications across industries deepen, the GenAI that was once shrouded in noise will gradually reveal its true nature. Recently, Hieb noted that businesses investing in GenAI do not necessarily have to choose expensive and complex solutions. In fact, more and more companies are discovering that large language models (LLMs) don't always...

Recently, Microsoft announced that its AI Vice President, Sebastian Bubeck, is leaving the company to join OpenAI. This marks a significant career shift for the seasoned employee who has worked at Microsoft for a full decade. Image source notice: the image is generated by AI, licensed from the service provider Midjourney. During his time at Microsoft, Bubeck was primarily responsible for developing small language models. Small language models are a technology aimed at balancing efficiency and effectiveness, with the goal of achieving performance similar to that of larger models without sacrificing performance.

Recently, NVIDIA launched the Minitron series of small language models, including 4B and 8B versions, significantly increasing training speed by 40 times while greatly reducing resource and data requirements, resulting in cost savings. By combining techniques of 'pruning' and 'knowledge distillation', the Minitron models maintain performance while reducing size, allowing developers to harness advanced technology for applications such as translation, sentiment analysis, and dialog AI at a lower cost. The open-source nature of the Minitron models enables more people to easily access and use them.

2023 has been one of the most disruptive years in the field of artificial intelligence, with a plethora of generative AI products entering the mainstream. Continuing its transformative journey, generative AI is expected to transition from an exciting topic to real-world applications in 2024. The field of generative AI is rapidly evolving, giving rise to a series of widespread trends that will facilitate the adoption of AI across various industries and its presence in our daily lives. Generative AI models are transcending text creation through the integration of multimodal functionalities.