在大会主论坛上，腾讯音乐娱乐集团法务管理部总经理岳魏捷围绕腾讯音乐在AI技术领域的突破与应用，发表了"以科技创新助力中国数字音乐高质量发 展"的主题分享。岳魏捷认为，人工智能正在重构音乐的生产、传播与消费方式，面对AI变革的浪潮，腾讯音乐始终积极拥抱数智时代，将AI技术渗透至音 乐产业链的各环节，并与多元应用场景深度融合。在音乐推荐与用户体验方面，腾讯音乐依托AI技术持续提升音乐发现的精准度与个性化程度，如旗下QQ 音乐成功将DeepSeek大模型集成到自研的AI助手中，显著提升了AI助手的理解能力，使智能交互更加精准全面，让用户更高效地触达心仪内容。 【环球网科技综合报道】9月21日消息，腾讯音乐娱乐集团（以下简称"腾讯音乐"或"TME"）亮相2025中国数字音乐产业大会，展示了其内容生态、前沿技 术、音乐公益等多元领域的代表性案例与重要成果。 创新传承传统文化是文化产业的重要方向，其中数字音乐扮演了重要角色。 在"数字音乐优秀传统文化传承"主题论坛上，腾讯音乐娱乐集团影视音乐总监、TME"影音+"&TME制作人联盟负责人李静楠以"用影视音乐传递中国传统文 化"为主题发表演讲，结合优秀案例深度探讨了影视音乐 ...

近日，恒为科技（603496）宣布停牌，筹划收购上海数珩信息科技股份有限公司（以下简称"上海数 珩"）的控股权，寻求AI战略的延伸。 事实上，作为A股上市公司中并购AI RaaS标的的首个案例，恒为科技更将之视作一场"卡位战"。恒为 科技高管在接受证券时报记者采访时表示，在AI下一轮竞争必然聚焦在应用和场景的大趋势下，恒为 科技此次发起并购，旨在以AI RaaS先行者的身份把握行业先机。 AI战略延伸 恒为科技9月16日晚间公告，正在筹划以发行股份及支付现金的方式购买上海数珩75%股份，同时拟发 行股份募集配套资金。本次交易预计不构成重大资产重组。 数珩信息从行业中脱颖而出的因素之一，或来自于对商业模式的创新——基于从基础模型到实际应用的 产业链全阶段能力，数珩信息正在实践AI RaaS的商业模式。这一尚不多见的商业模式被一些业内人士 视作AI应用落地的最终模式，有望应对AI软件、AI服务行业对企业技术整合能力、商业模式重构及风 险承担力的高要求。 成立于2017年的上海数珩于2022年率先开始训练可本地部署的S-GPT大模型，并在2023年3月上线完成 本地化部署。随后结合客户需求场景基于大模型自研出数智人 ...

Core Viewpoint - Qilu Bank, as the first local joint-stock commercial bank in Shandong Province, focuses on serving urban and rural residents, small and medium-sized enterprises, and local economic development through various financial services, including technology finance, green finance, inclusive finance, pension finance, and digital finance [1] Group 1: Technology Finance - Qilu Bank has developed a dual-direction mechanism between technology and finance, launching products like "KeChuang YiDai" to serve 61% of "specialized, refined, unique, and innovative" enterprises in Jinan [2] - The bank has established 2 technology branches, 3 specialty branches, and 10 technology finance centers to enhance the quality and efficiency of technology finance [2] - In 2024, the loan balance for technology enterprises reached 34.7 billion yuan, a 47% increase from the beginning of the year, continuing to grow to 38.8 billion yuan in Q1 2025 [3] Group 2: Green Finance - Qilu Bank prioritizes green finance in its annual credit policy, creating a product matrix that includes carbon assets and sustainable development-linked loans [4] - The bank's green loan balance reached 35.1 billion yuan in 2024, increasing by 10.9 billion yuan, and continued to grow to 42.2 billion yuan in Q1 2025 [5] Group 3: Inclusive Finance - Qilu Bank has launched the "Bearing Village Loan" to support the bearing industry in Linqing City, addressing funding gaps for over 5,400 related enterprises [6][7] - The bank has implemented a county financial development strategy, introducing over 80 specialized products, with inclusive loans accounting for over 20% of its total loans [7] Group 4: Pension Finance - The bank has developed a wealth management brand for pension finance, offering various products and achieving nearly 1 billion yuan in sales for its pension-related financial products [8] - Qilu Bank is actively involved in building a smart elderly care service system, integrating multiple resources to create a new model for elderly services [8] Group 5: Digital Finance - Qilu Bank is focusing on digital transformation by establishing a digital banking department and implementing various projects to enhance data processing efficiency and flexibility [9] - The bank is developing AI applications for various financial services, including intelligent contract review and customer analysis, to improve service delivery [9]

Core Insights - The article discusses a breakthrough in enzyme technology for plastic degradation, developed by a research team led by Professor Hong Liang from Shanghai Jiao Tong University, utilizing AI to design high-temperature resistant enzymes [1][4]. Group 1: AI and Protein Design - The research team has created an AI model named "Qimingxing" that constructs a functional map of proteins, enabling the efficient design of "super" proteins that are heat, acid, and alkali resistant [1][2]. - The AI model was trained using a dataset of 500 million proteins, tagged with functional labels based on environmental conditions such as temperature, pH, and pressure, making it the largest protein dataset globally [2][3]. Group 2: Efficiency and Automation - The team employs AI to select and modify protein templates to enhance their heat resistance, significantly improving the efficiency of protein design and testing compared to traditional high-throughput screening methods [3]. - An automated laboratory setup is in place to validate AI-designed proteins, creating a feedback loop that continuously optimizes the protein models, referred to as an "automatic driving mode" for proteins [3]. Group 3: Industry Impact - Eight industrial projects have utilized the "Qimingxing" model for protein design, achieving a success rate of 70%, with one biopharmaceutical company reporting a doubling of the lifespan of a designed protein, saving millions annually [3][4]. - The open-source release of the protein model and dataset on GitHub allows global research institutions to access and utilize this technology, potentially leading to a new industrial revolution in synthetic biology [4].