“AI向善”何以达成:深度学习与轻型应用——AI应用趋势下的新闻传播学思考  

How to Accomplish "AI Mileorisum":Deep Learning and Lightweight Applications

作　　者：张姣 曹轲[1] Zhang Jiao;Cao Ke

机构地区：[1]暨南大学新闻与传播学院,广东广州510630

出　　处：《传媒观察》2023年第7期28-34,共7页Media Observer

摘　　要：以ChatGPT为代表的新一代人工智能技术嵌入社会多个领域的趋势已经明朗,在新闻传播领域的轻型应用已经快速开始。本文对ChatGPT等AI技术应用趋势进行前瞻性分析后发现,技术的深度嵌入可能会产生以下五个问题:文本随机与深度伪造对新闻真实性构成挑战;私域流量与数据捆绑加剧被遗忘难题;用户生产造成信息淹没与新闻价值流动;社会超速与个人封闭召唤新闻预警功能;标准知识与认知偏差需要社会情境供给。The trend of the new generation of artificial intelligence technologies,represented by ChatGPT,being embedded in multiple fields of society has become clear,and lightweight applications in the field of journalism and communication have already begun rapidly.In this article,after a prospective analysis of the application trends of AI technologies such as ChatGPT,it was found that the deep integration of technology may raise the following five issues.Text randomness and deep fake challenge the authenticity of news.Private flow and data binding aggravate the problem of forgotten data.User production causes information overflow and the flow of news value.Social acceleration and individual closure call for news warning functions.Standard knowledge and cognitive bias require social context supply.

关 键 词：新闻真实 被遗忘权 信息生产 新闻预警 数据治理 

分 类 号：G206[文化科学—传播学] TP18[自动化与计算机技术—控制理论与控制工程]