研究表明，用户倾向于表现出一种 AI
排序理由 [lever_c_demoted from research: ic=1 ai=1.0]
在 Mastodon — mastodon.social 阅读 →
AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →
人们是懒惰地还是反思性地处理 # ChatGPT 推荐？三项随机实验操纵了 # AI 输出顺序、错误和声称的可靠性。
报道来源 [1]
-
Do people process # ChatGPT recommendations lazily or reflectively? Three randomized experiments manipulated # AI output order, errors, and alleged reliability.
Do people process # ChatGPT recommendations lazily or reflectively? Three randomized experiments manipulated # AI output order, errors, and alleged reliability. Users showed a “first‑is‑best” # bias — less so when warned about the unreliability of # LLMs . https:// doi.org/10.101…