智能算法推荐技术的逻辑理路、伦理问题及规制方略

Logic,Ethics and Regulation of Intelligent Recommendation Algorithme Technology

新技术的出现会给使用者带来新特权,这种特权往往会打破旧局面,挑战固有秩序,并引发一系列新问题。人工智能时代,"算法即权力"体现为经过用户和技术"赋权"后,算法推荐技术具有了筛选、定义、推送信息和记录、存储个人数据的特殊权力,并在某种意义上取代了传统媒体的价值判断和把关权。算法在拥有这些权力之后,其动机和执行结果并不都是"向善"的。在受众层面,信息茧房、数据滥用、算法黑箱、把关权的迁移等问题对用户的隐私权、自主权、知情权、平等权、被遗忘权造成了不同程度的威胁;在行业层面,新闻生产和传播方式发生变化,专业新闻机构的垄断性地位受到冲击;在社会层面,媒体的公共性及其作为社会公器的监督作用被削弱。面对算法推荐技术在新闻传播领域引发的伦理困境,应当坚持以人为本,重视人的价值,在数据采集、数据计算、算法模型中提升人的主导作用,寻找效率导向的"算法"与公平导向的"新闻"二者之间的平衡点,在此基础上建构算法推荐新闻的伦理规范原则及应用细则。

New technologies bring new privileges to their users.They often break old patterns and challenge previous orders,bringing about a series of new problems.In the era of artificial intelligence,"algorithm is power",which means that after being"empowered"by users and technology,recommendation algorithm technology has the special power to filter,define,deliver information and record,and store personal data,and thus in a sense replaces traditional media in value judgment and supervision.After getting these powers,algorithm is not always well-intentioned or does not always produce good results.In terms of audience,issues such as information cocoon rooms,data abuse,algorithmic black boxes,and transfer of supervision power pose threats of different degrees to users’privacy,autonomy,right to know,right of equality,and right to be forgotten.At the industry level,the way of news production and transmission has changed,and the monopoly status of professional news organizations has been impacted.At the social level,the publicity of the media and its supervisory role as a social public instrument have been weakened.Confronted with the ethical dilemma caused by recommendation algorithm technology in the field of journalism and communication,we should be people-centered,pay attention to human value,elevate the role of human being in data collection,data calculation,and algorithm modelling,and find a balance between efficiency-oriented"algorithms"and fairness-oriented"news",and on this basis make ethical principles and application regulations for algorithmic recommendation of news.