人工智能的快速发展已经对人类社会产生了深远的影响，且在众多领域得到应用。随着社会对AI能力需求的不断增长，现有AI技术的局限性日益凸显，这促使研究者和开发者向更高级别的智能——通用人工智能迈进。AGI在完成现实世界任务中的效率和取得的效果可与人类智能相媲美，这是AI发展的重要里程碑。目前，以算力和数据驱动的大语言模型和生成式AI成为迈向AGI的重要技术路径。“通用算法”向“通用模型”的转变意味着全新的研究和开发范式的形成。本报告重点介绍了大模型以及生成式AI这一新范式的发展概况、特点、面临的挑战、发展趋势，并提出相关建议。本报告旨在通过深入探讨AGI的现状和未来发展方向，促进研究人员和从业者对AGI的整体理解，并引发更广泛的关注与讨论。
The rapid development of Artificial Intelligence（AI）has profoundly impacted human society，driving significant progress in multiple fields.As the demands for AI capabilities continue to grow，the limitations of existing AI technologies become increasingly apparent，catalyzing the movement towards a higher level of intelligence—Artificial General Intelligence（AGI）. AGI’s efficiency and effectiveness in diverse real-world tasks can be comparable to human intelligence，reflecting a paramount milestone in AI evolution.Currently，Large Language Models（LLMs）or generative AI，driven by massive computational power and data，have become an important technical path towards AGI.The transition from “general algorithms”to “general models”represents a new and important paradigm for AGI research and development.This paper reviews the recent development，characteristics，challenges and trends of LLMs or generative AI.By exploring the current state and future trajectory of AGI，this paper aims to enhance a collective comprehension and catalyze broader discussions among researchers and practitioners on AGI.
