摘要
科研诚信技术治理是指在科研活动中以提高诚信水平、优化科研生态为目标,通过系统性地运用现代科技成果,对科研失信行为进行预防、识别和处置的治理活动。DeepSeek作为新一代人工智能,为推动科研诚信技术治理的多主体参与、准确性提升、全过程优化等提供了新契机。通过DeepSeek赋能,将有利于搭建高质量可共享的数据平台以推动协同性治理、完善智能检测与调查处理机制以推动精准化治理、夯实动态监测与信用评估体系以推动过程性治理,从而构建起科研诚信技术治理的人机协同实践路径。但也必须以有限赋能而非无限替代限定DeepSeek的赋能边界,始终坚持人类主权不可让渡的底线,及数据管理分析可信可控、技术系统动态优化、维护多元化协同治理等原则,提升科研诚信技术治理效能。
Technical governance refers to governance activities involving technical participation.Technical governance in research integrity is defined as technologically driven regulatory activities that systematically apply modern technological tools to enhance research integrity,optimize the scientifie ecogy8-tem,and combat research misconduct through prevention,identification,and resolution.As a frontier technology,DeepSeek-with its lightweight local deployment,Chain-of-Thought(CoT),and MuliModal Learning-offers new opprtunities for multi-agent participation,accuracy,and end-to-end optimization in research integrity governance.Looking ahead,DeepSeek can empower research integrity governance through three integrated approa-ches:building high-quality,shareable data platforms to advance collaboralive govemance;enhancing intelligent detection and investigation mechanisms to enable flexible govermance;and strengthening dynamic monitoring with credit assessment systems to drive process-oriented govermance.Collectively,these approaches will establish a human-AI collaborative framework for technical govemance in research integrity.Concurrently,vigilance against tech-nological hegemony necessitates constraining DeepSeek's role to supplementary empowerment-not substitution-while adhering to fundamental prinei-ples:inviolable human sovereignty,verifiable data control,adaptive systerm refinement,and diversified collaborative governance.This ensures optimal ef-ficacy in research integrity technical govemance.
作者
成尧
王健
CHENG Yao;WANG Jian(School of Marxism,Northeastern University,Shenyang 110167,China)
出处
《自然辩证法研究》
北大核心
2025年第8期137-144,共8页
Studies in Dialectics of Nature
基金
国家社会科学基金重大项目“习近平总书记关于科学技术重要论述的学理阐释”(24&ZD173)。
关键词
科研诚信
DeepSeek
技术治理
人机协同
research integrity
DeepSeek
technology govermance
Human-AI collaboration
作者简介
成尧(1995-),女,山西临汾人,东北大学马克思主义学院博士研究生,主要研究方向:技术伦理;王健(1967-),女,辽宁沈阳人,哲学博士,东北大学马克思主义学院教授,主要研究方向:技术伦理。
