如何匹配制造大省？江苏大模型期待“突围”

Core Insights - Jiangsu province has registered 28 large models, ranking fifth in the country, but there is significant room for improvement in both quantity and quality [1] - The development of large models in Jiangsu is seen as having great potential, especially in aligning with its manufacturing industry [1][10] Group 1: Current Status and Applications - The five newly registered large models target specific niches, including agricultural and intelligent manufacturing applications [1] - The "Qiancheng Cloud AI Intelligent Model" has been launched to enhance operational efficiency in rural e-commerce, providing services 1,000 to 2,000 times daily and supporting monthly orders of 100,000 yuan [2] - The "Changgan Model," the first cultural heritage model in the country, has been integrated into a digital museum platform, achieving over 35 million views and 850,000 registered users [3] - Jiangsu has over 100 typical application scenarios for large models across various sectors, including industry, finance, and healthcare, leading the nation [4] Group 2: Government Support and Infrastructure - Local governments and industrial parks are prioritizing large model enterprises and projects for development [5] - The "Suan Shu Space" in Nanjing aims to foster an AI innovation community, providing services for model registration and evaluation [6] - Suzhou has launched measures to support the development of large models, including financial incentives for models with over 100 billion parameters [8] Group 3: Challenges and Opportunities - The lack of talent who understand both large models and specific industries is a significant challenge for development [9] - Despite the challenges, Jiangsu's data resources and computing power are growing, with a total computing power of 33.8 EFlops, of which 19.4 EFlops is dedicated to large models, marking a 118% increase from last year [10] - The province's rich industrial data can support the training of vertical large models, and there is a call for stronger collaboration between academia and industry to cultivate talent [10][11]