大模型深度赋能行业形成AI+新质生产力

Empowering Industries with Large Models and Forming AI+New Quality Productive Forces

以GPT为代表的基座大模型快速进步,其认知能力的通用性、可完成任务的无限性以及生成内容的多样性等特点为行业深层智能化打开了全新的道路。然而大模型大规模落地于行业主体生产环节仍面临显著挑战。本文深入分析了大模型技术给产业智能化升级带来的机会以及形成“AI+新质生产力”所面临的挑战,并针对这些挑战提出了体系化人工智能技术体系,包括方法论以及相关的核心技术。

Over the past year,foundational models(FM)represented by GPT have made rapid progress.Their general abilities for understanding,performing unlimited tasks,and generating diverse contents open new directions for deep industrial intelligence.However,large-scale deployment of FM in primary industrial production processes still faces significant challenges.This paper analyzes the opportunities for industrial intelligence and the challenges in forming AI+new quality productive forces.To solve these challenges,we propose a new artificial intelligence technology system named Holistic Artificial Intelligence,which consists of overall architecture and core technologies.