Scholarlink

<p>The images used to depict artificial intelligence (AI) in science communication and marketing are often criticized for their stereotypical nature, yet their actual effectiveness remains an open question. Rather than reiterating existing critiques, this paper examines whether these images truly influence public perception and, if so, how different types of AI representations—particularly stock images—compare in their effectiveness. Drawing on Aristotelian rhetoric, we conceptualize effectiveness through three persuasive dimensions: <i>logos</i> (explanatory clarity), <i>ethos</i> (credibility and engagement), and <i>pathos</i> (emotional impact). To investigate these aspects, we conducted an experiment with 79 communication students, who were presented with three versions of the same fabricated scientific article: one without images, one with an institutional research image, and one with a stock image. The findings indicate that while stock images may reduce comprehension of technical content, they slightly enhance reader engagement. Our analysis calls for a more nuanced understanding of AI imagery, moving beyond ethical concerns to consider how these visuals shape public discourse. While stock images contribute to a standardized and sometimes misleading portrayal of AI, their rhetorical power in communication contexts should not be overlooked. The paper concludes by arguing for a more critical yet balanced approach to evaluating the role of images in AI representation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

How effective are depictions of AI? Reflections from an experimental study in science communication

  • 期刊论文

  • 发表日期：

    2025年3月7日

查看全文
  • Alberto Romele,
  • Marta Severo

摘要

The images used to depict artificial intelligence (AI) in science communication and marketing are often criticized for their stereotypical nature, yet their actual effectiveness remains an open question. Rather than reiterating existing critiques, this paper examines whether these images truly influence public perception and, if so, how different types of AI representations—particularly stock images—compare in their effectiveness. Drawing on Aristotelian rhetoric, we conceptualize effectiveness through three persuasive dimensions: logos (explanatory clarity), ethos (credibility and engagement), and pathos (emotional impact). To investigate these aspects, we conducted an experiment with 79 communication students, who were presented with three versions of the same fabricated scientific article: one without images, one with an institutional research image, and one with a stock image. The findings indicate that while stock images may reduce comprehension of technical content, they slightly enhance reader engagement. Our analysis calls for a more nuanced understanding of AI imagery, moving beyond ethical concerns to consider how these visuals shape public discourse. While stock images contribute to a standardized and sometimes misleading portrayal of AI, their rhetorical power in communication contexts should not be overlooked. The paper concludes by arguing for a more critical yet balanced approach to evaluating the role of images in AI representation.