Gartner预测到2026年60%的AI项目将因缺乏这4大架构支柱而失败
gartner预测到2026年由于数据准备不足，60%的ai项目将被废弃，强调可靠的数据管道和上下文工程对于从原型转向可扩展的生产级ai系统至关重要。
gartner predicts that organizations will abandon 60% of ai projects by 2026 if they lack ai-ready data, according to an mit technology review insights report in partnership with elastic. the report identifies four enduring architectural pillars necessary for scaling ai: data quality, context engineering, governance, and human expertise. it emphasizes that while models evolve, the foundation must support reliable retrieval and decision-making, highlighting that context engineering—often more critical than prompt engineering—ensures models receive precise, relevant inputs to maintain speed and accuracy. successful architectures require real-time, unified data foundations and vector databases to support advanced techniques like retrieval augmented generation (rag) as companies move toward agentic systems.