算法之下:数字信任还是专家信任?  

Under the Algorithm:Digital Trust or Expert Trust?

作　　者：吴新慧[1] Wu Xinhui

机构地区：[1]杭州电子科技大学法学院,杭州310018

出　　处：《复印报刊资料（社会学）》2023年第4期104-111,共8页SOCIOLOGY

基　　金：教育部人文社会科学研究青年基金项目“新生代农民工数字融入与市民化研究”(项目编号:19YJC840045)。

摘　　要：随着算法的广泛应用,公众在决策过程中多了一种选择:可以信任专家,也可以信任算法。算法是在软件中提供控制加逻辑的工具,具有不透明性、不确定性和权威性等特征。算法自主性程度、任务风险大小以及人类和算法的交互程度影响着公众选择信任算法还是专家。媒介算法中的内容生产使专家识别变得困难,面对众多“专家意见”,人们无所适从;搜索、推荐算法带来的“以自我为中心的偏见”和“回声室”效应,使个人更加固执已见;咨询算法、人机虚拟团队提高了专家能力,但不必然带来公众对专家的更高评价;执行算法则加深了公众对专家的依赖。当前的智能社会,公众需要认识专家地位及其认知局限性,不盲从、不苛责,更需要提升自身算法素养,校准人机信任。With the widespread adoption of algorithms,the public has one more choice in the decision-making:they can trust the experts or the algorithm.Algorithms are tools that provide control plus logic in software,which are characterized by opacity,uncertainty,and authority.Algorithmic autonomy,task risk,and human-algorithmic interaction influence public trust choices.Content production algorithms make expert identification more difficult.The public doesn't know who to trust.The search algorithm reinforces"self-centered bias"and the recommendation algorithm makes the"echo chamber".Both of them make individuals more stubborn.Consulting algorithms and humanmachine virtual teams improve the capabilities of human experts,but they do not necessarily lead to higher public evaluation of experts.Executing algorithms make the public more dependent on experts.In the intelligent society the public needs to understand experts'status and their cognitive limitations,and should not blindly follow or criticize experts.It needs to calibrate human-machine trust by cultivating public algorithm literacy.

关 键 词：算法 算法信任 专家信任 智能社会 

分 类 号：C912.6[经济管理] TP301.6[社会学]