Investment Rating - The report maintains a positive outlook on the industry, specifically highlighting the potential for effective control of AI model hallucinations by 2026 [2]. Core Insights - The report emphasizes that while hallucinations in AI models are inevitable, advancements in algorithms, data quality, and engineering practices can significantly reduce their occurrence. The top 25 global models have achieved a hallucination rate below 8% [5][6]. - The report identifies three key areas for investment: mature AI applications, marketing AI that is less sensitive to hallucinations, and data plus AI infrastructure [6]. Summary by Sections 1. Hallucinations - The Lower Bound of Model Capability - The report defines hallucinations as overconfident errors produced by language models, which can include fabrications, factual inaccuracies, contextual misunderstandings, and logical fallacies. For instance, GPT-3.5 had a hallucination rate of approximately 40%, while GPT-4's rate was 28.6% [14][15]. 2. Sources of Hallucinations - Hallucinations arise from several factors, including model architecture, toxic data, lack of accuracy in reward objectives, and context window limitations. Addressing these factors is crucial for controlling hallucinations [7][8]. 3. Reducing Hallucinations: From Models, Data, Engineering, and Agents - The report discusses various strategies to mitigate hallucinations, such as using larger training datasets, extending context windows, and incorporating human feedback through reinforcement learning (RLHF) [25][26]. - Engineering practices like Retrieval-Augmented Generation (RAG) are becoming standard, with Gartner predicting a 68% adoption rate by 2025 [56][57]. 4. 2B Application Penetration and Evolution - The report notes that the control of hallucinations in mainstream models has made significant progress, with the top 25 models in the Vectara HHEM ranking achieving hallucination rates below 8%. For example, the Finix model developed by Ant Group has a hallucination rate of only 1.8% [72].

Core Insights - The report highlights the limitations of traditional Business Intelligence (BI) systems, which struggle to meet the demands for real-time and dynamic decision-making due to their closed architectures and static processing capabilities [1][21][24] - The integration of Artificial Intelligence (AI) with BI, termed Artificial Intelligence and Business Intelligence (ABI), is driving a shift from reactive to proactive decision-making, with ABI expected to experience explosive growth in China, reaching a market size of 800 million yuan in 2024 and a CAGR of 42% from 2024 to 2028 [1][11][13] - Key drivers for ABI growth include deepening enterprise reliance on data, breakthroughs in AI technology, and supportive policies [1][11] Industry Overview - ABI leverages technologies such as Natural Language Processing (NLP) and machine learning to enable conversational interactions, multimodal data analysis, and complex reasoning, enhancing decision-making across various sectors including finance, retail, manufacturing, government, and energy [2][3] - The financial sector utilizes ABI for intelligent risk control and quantitative trading, while retail benefits from dynamic pricing and inventory optimization [2][3] - Manufacturing employs predictive maintenance and process optimization to reduce downtime, and government sectors enhance service efficiency through smart traffic and urban governance [2][3] Market Dynamics - The ABI market in China is projected to grow from 300 million yuan in 2023 to 800 million yuan in 2024, driven by the increasing complexity of decision-making needs and the inadequacies of traditional BI tools [1][11][13] - ABI's core challenges include data governance lag, algorithm opacity, fragmented scenarios, and high technical costs, with future trends focusing on edge computing, real-time analysis, generative AI penetration, and privacy computing technologies [3][11] Technological Advancements - ABI employs advanced techniques such as Text2SQL and Text2DSL to convert natural language into data queries, enhancing the depth of analysis through external knowledge integration and multi-agent collaboration [2][3][30] - The integration of AI allows for the automation of data processing, significantly improving efficiency and enabling strategic decision-making by providing deeper insights and optimizing resource allocation [40][42] Future Outlook - The ABI landscape is evolving towards democratization and intelligence, reshaping the decision-making paradigm driven by data within enterprises [3][11] - Major global players like Microsoft and Salesforce focus on ecosystem integration, while domestic firms like Alibaba Cloud and Fanruan emphasize lightweight deployment and localized innovation [3][11]

Investment Rating - The report maintains a "Hold" rating for the industry [6] Core Insights - The development of AI Agents is driving a transformation in software forms, establishing databases as indispensable in the AI era, serving not only as data carriers but also helping to mitigate issues like hallucinations in large model reasoning [5][42] - AI is empowering databases to upgrade themselves, enhancing operational efficiency and driving industry growth [4][31] Summary by Sections DB for AI: AI Agents Driving Software Transformation - AI Agents are expected to interact directly with databases, potentially replacing the intermediary application layer in traditional software architectures [1][11] - Databases play a crucial role in the AI era by ensuring high-quality data for AI training, which is essential for effective AI model performance [2][14] - Technologies like vector databases and RAG (Retrieval-Augmented Generation) are directly empowering AI development, addressing issues such as hallucinations in large model reasoning [2][16] AI for DB: AI Empowering Database Upgrades - Intelligent operations are being implemented, allowing for real-time monitoring, predictive analysis, and automated processing of database systems [4][31] - The use of natural language processing enables users to interact with databases more easily, converting natural language into SQL queries [4][35] - Autonomous databases are emerging, utilizing machine learning to perform tasks traditionally handled by database administrators, such as optimization and maintenance [4][36] Investment Recommendations - The report suggests focusing on companies such as Dameng Data, Taiji Co., Haima Data, Softcom Power, Creative Information, Star Ring Technology, SuperMap Software, and Toris [5][42]