Group 1 - The data summit held in Hong Kong on July 28 focused on optimizing trade financing processes and facilitating lending for SMEs through data infrastructure [1] - The Hong Kong Monetary Authority (HKMA) shared developments on the "Commercial Data Interchange," emphasizing four key areas: CargoX trade financing project, government data openness, SME credit data analysis, and cross-border data verification [1] - HKMA President Eddie Yue highlighted the authority's commitment to enhancing Hong Kong's data infrastructure to accelerate operational processes for financial institutions and SMEs [1] Group 2 - The Secretary for Transport and Logistics of Hong Kong mentioned efforts to build a port community system that connects all stakeholders in the cargo transport process, ensuring transparency in cargo status [1] - The Hong Kong International Airport is recognized as the world's largest cargo airport, and the Airport Authority is leveraging technology to enhance competitiveness through digitalization of the air cargo supply chain [2] - The Hong Kong Banking Association encourages member banks to participate in the "Commercial Data Interchange" projects to expedite loan approval processes and improve lending risk management for SMEs [2]

Core Insights - The Hong Kong Monetary Authority (HKMA) is committed to enhancing Hong Kong's data infrastructure and ecosystem to facilitate trade financing and support small and medium-sized enterprises (SMEs) [1][2][3] Group 1: Data Infrastructure and Trade Financing - The "Commercial Data Hub" aims to simplify trade financing processes and enhance innovation for SMEs, thereby driving economic growth [1][2] - The Cargox project is being developed to improve the digital ecosystem for trade financing, with pilot banks testing the verification of trade authenticity using logistics data [2][3] - The HKMA is collaborating with the Hong Kong government to connect the "Commercial Data Hub" with the upcoming "Single Trade Window" service, allowing users to share customs data with banks to expedite loan approvals [2][3] Group 2: Credit Data Analysis and Cross-Border Verification - The HKMA is working on a proof-of-concept for "Commercial Credit Database 2.0" to create a credit scoring model for SMEs, which is expected to simplify loan applications and reduce borrowing costs [2][3] - The "Commercial Data Hub" has successfully integrated with the Shenzhen-Hong Kong cross-border data verification platform, enabling banks to process personal and corporate loans more efficiently [3] Group 3: Industry Support and Future Directions - The Hong Kong International Airport is leveraging advanced technologies like blockchain to enhance cargo supply chain efficiency and support SMEs in financing processes [4] - The Hong Kong Banking Association is encouraging member banks to participate in projects related to the "Commercial Data Hub" to accelerate loan approvals and improve risk management for SMEs [5]

当主流大模型的参数规模皆突破万亿级，行业逐渐意识到一个残酷现实：大模型的技术红利窗口正在收 窄。 Transformer架构主导的算法同质化，使得模型性能差异从技术代差演变为工程优化。与参数量的边际效 益持续递减形成鲜明对比的是，IDC研究显示企业80%的非结构化数据仍处于沉睡状态，包括设备日 志、工艺文档、客户对话等数据金矿普遍尚未被有效开采。 这种背景下，企业打造AI能力的重心正从"模型军备竞赛"转向"数据基建深耕"。日前，AI基础软件设施 供应商星环科技发布的AI-Ready Data Platform通过技术架构的重构，正在加速这一新局面的打开。 一、AI竞争聚焦数据，传统平台与需求矛盾重重 Gartner高级研究总监方琦的观察直指行业痛点：由于大多数企业都依赖于相似的预训练模型，因此具 有企业自身特色的数据正成为生成式人工智能采用和创新的关键差异化因素。企业独特数据，包括私有 数据和行业Know-how，成为AI竞争的唯一壁垒。 然而，传统数据平台在数据治理、整合和管理方面存在诸多不足，与大模型对高质量、多模态、领域知 识数据的需求形成了核心矛盾。 数据存储的割裂问题成为了显著的瓶颈。企业通常需要管 ...