VeriSilicon President Predicts Consolidation of China's AI Models by 2028
During the World Artificial Intelligence Conference 2024 in Shanghai, Dai Weimin, President of VeriSilicon, addressed the RISC-V and Generative AI Forum with a presentation on the opportunities and challenges facing AIGC chips. He highlighted that achieving human-level intelligence requires exponentially increasing computational power and model parameters. Dai criticized the current global race to develop proprietary large-scale AI models, describing it as a chaotic competition that results in significant energy waste. He noted that over 100 such models currently exist in the Chinese market alone. Looking ahead, Dai predicted a major consolidation in the industry, stating that by 2028, China will likely have fewer than ten foundational large-scale models, with an ideal target of just five. This forecast suggests a shift from fragmented development to a more streamlined approach focused on efficiency and core foundational technologies, aiming to reduce resource consumption while maintaining technological advancement in the artificial intelligence sector.
Wire timeline
VeriSilicon President Predicts Consolidation of China's AI Models by 2028
During the World Artificial Intelligence Conference 2024 in Shanghai, Dai Weimin, President of VeriSilicon, addressed the RISC-V and Generative AI Forum with a presentation on the opportunities and challenges facing AIGC chips. He highlighted that achieving human-level intelligence requires exponentially increasing computational power and model parameters. Dai criticized the current global race to develop proprietary large-scale AI models, describing it as a chaotic competition that results in significant energy waste. He noted that over 100 such models currently exist in the Chinese market alone. Looking ahead, Dai predicted a major consolidation in the industry, stating that by 2028, China will likely have fewer than ten foundational large-scale models, with an ideal target of just five. This forecast suggests a shift from fragmented development to a more streamlined approach focused on efficiency and core foundational technologies, aiming to reduce resource consumption while maintaining technological advancement in the artificial intelligence sector.
TechNode