2026科技棱镜年度报告_聚焦技术驱动的业务变革_46页_11mb

Looking Glass Summary

Core Content

The Looking Glass report by Thoughtworks provides a comprehensive overview of how AI is reshaping the technology landscape across enterprises. It emphasizes that the transformation is not just about adopting AI as a standalone tool, but about rethinking how technology is built, operated, and governed to support a more adaptive, intelligent, and responsible future.

The report outlines five key lenses through which the impact of AI can be understood: AI and software delivery, preparing for agentic transformation, evolving interactions, data platforms, and responsible AI foundations. Each lens explores how AI is being integrated into different aspects of the enterprise, from development to operations and customer engagement.

Main Points

1. AI and Software Delivery

• AI-First Software Delivery (AIFSD): This involves integrating generative and agentic systems into the full software development lifecycle, enhancing efficiency and quality.
• Goal-Based Development Environments (GBDEs): Developers can now specify high-level objectives, and AI agents handle the implementation details, reducing the gap between business intent and software creation.
• Continuous Learning Delivery Systems: Feedback from users and telemetry is used to continuously improve software through neural delivery loops.
• Neural Software Twins: These are living models of systems that allow AI to experiment, predict, and suggest changes before they are implemented in production.
• Synthetic Engineers: Composite AI entities that can manage development streams with minimal human intervention, hinting at the future of collective AI design teams.
• Multimodal Collaboration: AI facilitates real-time translation and multimodal reasoning, enabling seamless global collaboration.

2. Preparing for Agentic Transformation

• The shift from isolated AI experiments to agentic transformation is inevitable. The focus is on enabling AI agents to operate autonomously across workflows, reducing human bottlenecks.
• Agentic Workflows: AI agents are managing complex, multi-step business processes, improving responsiveness and productivity.
• Embedded AI Governance: New governance models are integrating transparency and accountability directly into AI operations.
• AI-Enhanced Decision Systems: AI is moving from descriptive analytics to co-decision systems, where it continuously augments human judgment.
• Data Fabrics and Synthetic Data Ecosystems: These are unlocking value from sensitive or limited datasets, addressing the data bottleneck for AI.
• Human-AI Co-Decision Systems: These systems combine human judgment with AI predictions to make more informed and impactful decisions.

Key Information

Opportunities

• Break through modernization barriers: AI tools like CodeConcise can significantly reduce the time and effort needed to modernize legacy systems.
• Elevate the way teams build: AI can speed up software delivery and enable faster, more intelligent product prototyping and design.
• Lower operational risks and enhance reliability: AIOps is improving incident detection and resolution times, enhancing system stability.
• Enhance developer satisfaction: By offloading repetitive tasks, AI allows developers to focus on higher-value work and innovation.

Case Study: Automotive Manufacturer

A leading automotive manufacturer faced challenges in modernizing a legacy system with millions of lines of code. Using the CodeConcise tool, the time to reverse engineer 10,000 lines of code was reduced by two-thirds, saving up to 60,000 person-days and boosting confidence in modernization.

Case Study: Global Pharmaceutical Company

A global pharmaceutical company used AI to transform its engagement with healthcare professionals. An AI-powered next-best-action engine and contextual assistant enabled more personalized, efficient, and insightful interactions, improving both effectiveness and trust.

Trends to Watch

Adopt (Current Trends)

• Generative AI for enterprise
• Retrieval-augmented generation (RAG)
• AI code assistants
• MLOps and model operationalization
• AI governance and responsible AI
• AIOps for IT operations
• AI-ready data infrastructure
• Unified observability platforms
• Configuration management/policy-as-code
• Voice user interfaces
• Cloud-native application protection
• DataOps and data engineering
• iPaaS with AI capabilities
• Low-code workflow automation
• Real-time translation
• AI-native platforms
• AI-powered predictive maintenance
• Agentic conversational AI

Analyze (Emerging Trends)

• Data mesh architecture
• Composable enterprise/PBCs
• API marketplaces and ecosystems
• Distributed cloud infrastructure
• Event-driven architecture
• Digital twins for enterprise
• AI-powered drug discovery
• Additive manufacturing
• Private 5G networks
• Autonomous mobile robots
• Domain-specific AI models
• AI for full SDLC
• Sustainable/Green Cloud Computing
• Data products/data as a service
• Small language models
• Multimodal AI
• Industry cloud platforms
• AI-powered threat detection
• Continuous threat exposure management
• Semantic layer/metrics stores
• AI-embedded enterprise applications
• Headless commerce
• Event-Driven Integration
• Autonomous vehicles
• Intent-based networking
• LEO satellite connectivity
• Digital Thread in Manufacturing
• GenAI virtual assistants

Anticipate (Future Trends)

• Post-quantum cryptography
• AI-powered cloud management
• AI agents for data and analytics
• Total experience (TX) platforms
• Neuromorphic computing
• Advanced materials (metamaterials)
• AI-driven capacity planning
• Emotion AI/affective computing
• Artificial general intelligence (AGI)
• 6G Research
• General-purpose humanoid robots
• Machine customers/autonomous commerce
• Full autonomous vehicles
• Programmable money/CBDCs
• Industrial metaverse

Actionable Advice

Things to Do (Adopt)

• Integrate AI into talent strategies: Teams need to adapt to new workflows and understand the broader goals AI serves.
• Define and track AI ROI: Establishing clear value definitions and measuring progress is crucial for the success of AI initiatives.
• Ensure transparency and human oversight: In co-decision systems, AI reasoning, evidence, and data sources must be inspectable and verifiable.

Things to Consider (Analyze)

• Multi-agent systems: As AI agents become more prevalent, businesses should consider how multiple agents can collaborate to handle complex tasks.
• Tighter AI regulation: Increasing global scrutiny on AI usage is driving the need for more robust compliance and governance frameworks.

Conclusion

The Looking Glass report highlights that AI is not a singular disruptive force but a transformative element that is redefining how enterprises build and operate their technology estates. The focus is on creating adaptive, responsible, and AI-ready systems that support both human and machine collaboration. Organizations that proactively adopt and analyze these trends will be better positioned to navigate the future of technology and leverage AI to drive innovation, efficiency, and customer-centric outcomes.