Generative AI Models

In a blog post today (October 21), Samsung announced the launch of its second-generation generative AI model, Samsung Gauss2, at a developer conference in South Korea. The multimodal language model is capable of processing multiple data types such as text, code, and images simultaneously, providing a significant boost in performance and efficiency. Gauss2 is available in three different sizes - Compact, Balanced and Supreme - to meet the needs of different computational requirements. Gauss2 offers three different models with different specifications to meet the needs of different computing environments and application scenarios, which are briefly summarized by IT Home as follows...

According to foreign media reports, the AI image generator Midjourney quietly revised its terms of service last week, replacing the previous humorous words with more formal lawyer language. The change is seen as a signal that Midjourney is confident that it can win a potential copyright lawsuit with creators. Source note: The image is generated by AI, and the image licensing service provider Midjourney is a generative AI model. Midjourney uses a large number of existing images and texts for training, including a large number of copyrighted works. The company has always argued that according to the "…

Canalys released a forecast of the top ten trends in the global technology industry in 2024. These include the popularity of generative AI, the shortage of AI talents, the increase in corporate supercomputers, the increase in personal code generation through informal programming training, the shortage of Apple Vision Pro products, the impact of national control on generative AI models, and the resumption of growth in US channels from April 2024. The forecast shows that these trends will have a profound impact on the global technology industry. The following are the specific contents: 1. By 2026, most software and hardware products will have generative AI…