2023-10-18-Gartner-生成式+AI+将如何改变您的行业_47页_3mb

Gartner Panel Discussion Summary

Core Content

This document outlines Gartner's insights on how Generative AI (GenAI) is transforming various industries, including Banking, Healthcare, and Manufacturing. It provides a structured view of the potential use cases, the associated value and feasibility of these applications, and the regulatory considerations. The content is presented through Use-Case Prisms and Scorecards, which assess the impact and readiness of GenAI across different business functions and sectors.

Main Points

Overview of Generative AI

• Generative AI is a subset of AI that creates new, original artifacts based on data and models.
• Large Language Models (LLMs) are a type of foundation model focused on natural language processing.
• Foundation Models are trained on vast, unlabeled datasets and can be fine-tuned for specific applications.
• ChatGPT is an example of an LLM that generates content based on a foundational model and reinforcement learning.

Industry Use Cases and Scorecards

• Gartner has developed Use-Case Prisms to evaluate the impact of GenAI across multiple industries.
• These prisms are divided into value (Increase Revenue, Increase Efficiency, Manage Risk, Nonfinancial Value) and feasibility (Technical Feasibility, Internal Readiness, External Readiness) dimensions.

Banking Industry

• Current and Planned Usage: Half of banking CIOs have adopted or are planning to deploy GenAI.
• Key Use Cases:
  • Fraud Prevention: 13% currently implementing, 13% planning to implement within the next year.
  • Code Generation & Conversion: 12% currently implementing, 18% planning to implement.
  • Contact Center Assistant: 8% currently implementing, 17% planning to implement.
  • AML & Compliance: 6% currently implementing, 9% planning to implement.
  • Personalized Product Recommendation: 5% currently implementing, 14% planning to implement.
• Examples:
  • Swedbank uses GANs to improve fraud and money-laundering prevention.
  • Stripe uses LLMs to route tickets, summarize user questions, and assist with fraud detection.
  • Morgan Stanley leverages GPT to deliver insights to financial advisors and streamline client communications.
  • Commonwealth Bank of Australia uses ChatGPT to transcribe and summarize customer service calls, and to assist with complex customer queries.

Healthcare Providers

• Use Cases:
  • Autogenerating Clinical Documentation
  • Autonomous Clinical Coding
  • Healthcare Consumer Language Translation
  • Autonomous Virtual Primary Care
• Scorecard Highlights:
  • High scores for efficiency and nonfinancial value in most use cases.
  • Autonomous Virtual Primary Care scores highest in value (4) and efficiency (4).

U.S. Healthcare Payers

• Use Cases:
  • Digital Broker
  • First-Draft Member Communications
  • Member EOB and Provider Bill Matching Tool
  • Automated Policy Administration
  • Autonomous Underwriting and Actuarial
• Scorecard Highlights:
  • High feasibility for nonfinancial value and increased efficiency.
  • Autonomous Underwriting and Actuarial scores high in value (3.1) and feasibility (2.5).

Manufacturing

• Use Cases:
  • Data Insights
  • AI Image Generation
  • Materials Processing Efficiency
  • Mechanical Design Optimization
  • Supply Chain Optimization
• Scorecard Highlights:
  • High feasibility for increased efficiency and managed risk.
  • Supply Chain Optimization scores high in value (3.5) and efficiency (3.5).

Key Information

Industry Adoption

• Banking: 50% of CIOs have adopted or are planning to deploy GenAI.
• Healthcare: Gartner has identified numerous use cases for GenAI, with a focus on improving documentation, communication, and administrative tasks.
• Manufacturing: Use cases include design optimization, quality intelligence, and workforce training.

Risks and Challenges

• Unreliable Outputs: LLMs may produce faulty or inaccurate information due to training data limitations.
• Data Privacy and Confidentiality: Sensitive information may be incorporated into training data.
• Intellectual Property (IP) Concerns: Generated content may infringe on existing IP.
• Liability: Uncertainty around liability for AI outputs remains.
• Cybersecurity: Potential for misuse in generating malware or accessing data at scale.
• Consumer Protection: Failure to disclose AI usage may erode customer trust.
• Regulatory Compliance: The AI regulatory landscape is fragmented and fast-paced, making compliance challenging.

Regulatory Scenarios

• Innovation-First: Minimal oversight, focus on ethical standards.
• Harm Reduction-First: Regulatory frameworks with caution and penalties.
• Surveillance-First: Tight monitoring and high penalties for misuse.

Tech Roles and Leadership

• Tech Roles are leading the development of GenAI strategies in the financial services sector.
• Roles Leading the Efforts: Tech roles are the primary drivers of GenAI initiatives.

Conclusion

Gartner's report highlights the transformative potential of Generative AI across Banking, Healthcare, and Manufacturing, emphasizing the importance of value and feasibility in evaluating its impact. It also outlines the risks and regulatory challenges that organizations must address when implementing GenAI solutions. The report serves as a comprehensive guide for businesses looking to adopt or plan for the use of Generative AI in their operations.