高级检索

人机智能协同综述:哲学、技术和应用视角

A Overview on Human-Computer Synergy: Philosophical, Technical, and Application Perspectives

  • 摘要: 协同指两个及以上的个体或资源协调工作,以完成特定任务的过程与能力,具体表现为人人、机机或人机的协同。早期的人机协同主要为交互。然而,随着人工智能技术的迅猛发展,人机关系已演变为智能伙伴关系,人机共生时代已然来临,智能协同成为必然。本文从哲学、技术和应用3个维度对人机智能协同展开探讨。从哲学维度,梳理人机及其智能能力的本体论、认识论和实践论等相关议题,并通过剖析人人协同,明确了人的意愿对协同增效的作用及增强意愿的“诱因”。从技术维度,人机协同经历了3条主要演进路径:人机交互从人机沟通发展到多模交互,再发展到人机互动;控制决策从单纯自动控制迈向人机协同控制;人工智能系统从简单模拟智能到拓展人类思维和认知能力。从应用维度,介绍了人机协同在智能家居、智慧教育和智慧医疗领域的应用,并讨论了生成式、认知式和嵌入式人工智能适合的应用场景。最后,探讨了人工智能面临的幻觉与伦理问题,分析了问题成因,并指出了人工智能在发展过程中需深思的若干未决问题。

     

    Abstract: Synergy refers to the process and capability of at least two distinct individuals or resources working in concert to accomplish specific tasks. It manifests as synergy between humans, between computers, and between human and computer. Early human-computer collaboration primarily existed in interactive forms. With the rapid advancement of AI technology, human-computer relationships have gradually evolved into intelligent partnerships, ushering in an era of symbiosis between humans and computer, where intelligent synergy has become inevitable trend. This paper investigates human-computer synergy through three dimensions: philosophy, technology, and application. From the perspective of philosophy, it summarizes the ontological, cognitive and practical problems of human-computer or human-AI system. Through analyzing human-human collaboration, it clarifies how human intentions enhance synergistic effects and identifies the “motivators” that reinforce these intentions. From a technical perspective, human-computer synergy is undergoing three major evolutionary pathways: human-computer interaction has progressed from basic communication to multi-modal interaction and ultimately to human-computer synergy; control decision-making has shifted from purely automated control to collaborative human-computer control; AI systems have evolved from simple simulation of human intelligence to expanding human cognitive capabilities. From an application perspective, this paper introduces practical implementations of human-computer synergy in smart home, smart education, and smart healthcare domains, while discussing suitable application scenarios for generative AI, cognitive AI, and embedded AI. Finally, the article discusses the urgent hallucination and ethical issues facing current AI, analyzing their primary causes: insufficient model stability and lack of data purity. Additionally, the article proposes unresolved issues that require deeper consideration for the future development of AI.

     

/

返回文章
返回