English(EN) If I were to suggest one small tweak, I’d say *undesired* (rather than undesirable) output, to highlight the relativity to user expectations and avoid the value

AI术语辩论：“undesired output”与“agentic AI”

作者 PulseAugur 编辑部 · [1 个来源] · 2026-07-04 10:15

关于AI术语的一个建议是使用“undesired output”（不期望的输出）而不是“undesirable output”（不良输出），以强调输出是相对于用户期望而言的，而不是对其固有质量的判断。术语“agentic AI”也被讨论为“probabilistic automation”（概率性自动化）的一个变体。 AI

影响通过精确的语言，改进对AI输出和agentic（自主的）能力的理解。

排序理由该条目讨论了与AI概念相关的细微术语和措辞，符合“评论”类别。

在 Mastodon — fosstodon.org 阅读 →

观点

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

报道来源 [1]

Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] · 2026-07-04 10:15

If I were to suggest one small tweak, I’d say *undesired* (rather than undesirable) output, to highlight the relativity to user expectations and avoid the value

If I were to suggest one small tweak, I’d say *undesired* (rather than undesirable) output, to highlight the relativity to user expectations and avoid the value judgement. “Agentic AI” is more tricky, but that’s really just a variant of probabilistic automation, the authors’ pref…

报道来源 [1]

If I were to suggest one small tweak, I’d say *undesired* (rather than undesirable) output, to highlight the relativity to user expectations and avoid the value

相关实体

相关话题

If I were to suggest one small tweak, I’d say undesired (rather than undesirable) output, to highlight the relativity to user expectations and avoid the value