遇见数据集

Theatrical AI Agency Index (TAAI) Dataset: YouTube Audience Discourse on AI as Creative Agent in Theatre and Performance (n=2,816, April 2026)

收藏
Zenodo2026-04-26 更新2026-05-26 收录
官方服务:

资源简介:

This dataset contains 2,816 YouTube comments collected viathe YouTube Data API v3 (April 2026) across five theatre-specific discourse domains targeting naturalistic audienceresponses to AI as a creative agent in live performance: (1) AI written play theatre performance audience reaction 2025(2) artificial intelligence actor robot performance stage 2025(3) AI generated theatre script play human creativity 2025(4) can AI replace human actors theatre performance debate 2025(5) AI theatre performance real or fake authentic art 2025 The corpus is the primary empirical dataset for theTheatrical AI Agency Index (TAAI) — the first computationalmetric measuring whether theatre audiences accept or refuseAI as a legitimate creative agent in live performance,providing audience-reception data to complement practitioner-focused AI theatre research. TAAI = Theatrical Agency Acceptance / (TAA + Agency Refusal)1.0 = pure Agency Acceptance0.5 = contested zone0.0 = pure Agency Refusal Three signal types grounded in performance theory: THEATRICAL AGENCY ACCEPTANCE (TAA): TAAI numerator → "AI can write plays", "AI as co-creator", "impressive AI performance", "AI brings new possibilities" Grounded in: Barad (2007) posthuman agency; Fischer- Lichte (2008) transformative performance theory AGENCY REFUSAL (AR): TAAI denominator partner → "AI has no soul", "not real theatre", "can't replace human actors", "fake art" Grounded in: Benjamin (1935) aura; Aristotle Poetics mimesis; Auslander (2008) liveness ONTOLOGICAL ANXIETY (OA): novel modifier signal → "what does this mean for theatre?", "who is the author?" "is this still art?", philosophical uncertainty Grounded in: Sontag (1966); Benjamin (1935) aura dissolution; distinct from ASAI/AEDI fear signals KEY FINDINGS:- Total corpus: n=2,816 comments, ~50 videos- Categorised (TAAI computed): 10 (0.4%)- Mean TAAI: 0.6000 — TAA marginally dominant- TAA cluster: 6 (60% of categorised)- AR cluster: 4 (40%)- Ontological Anxiety: 1 (novel signal confirmed)- TAA density: 0.0129/100 tokens- AR density: 0.0089/100 tokens- OA density: 0.0002/100 tokens (rare but present) DOMAIN-LEVEL TAAI:- "real or fake authentic art": TAAI=0.7500 (TAA dominant)- "AI written play / audience reaction": TAAI=0.5000- "AI generated theatre / human creativity": TAAI=0.5000 CRITICAL FINDINGS:Most-liked TAA comment (2,713 likes): "Why does it feellike the AI actors never met eye to eye" — acceptancecombined with uncanny valley perception, instantiatingFischer-Lichte's (2008) co-presence theory.Most-liked AR comment (1,842 likes): "Someone said weshould call AI people 'Fartists' since it's fake art"— collective ironic Agency Refusal confirming Benjamin's(1935) aura theory in naturalistic discourse.OA signal confirmed: "As someone who's made a careerreimagining the Great American Songbook, if the AI[continues...]" (214 likes) — philosophical authorshipanxiety, first computational evidence of OA in theatreaudience discourse. Files:- taai_videos.csv: ~50 unique videos metadata- taai_comments.csv: 2,816 raw comments- taai_results.csv: TAAI scores + signal annotation Method: YouTube Data API v3. Python 3.12, April 2026.

本数据集通过YouTube Data API v3(2026年4月)收集了2816条YouTube评论,覆盖5个聚焦戏剧主题的讨论领域,旨在探究观众对AI作为现场表演中创造性AI智能体(AI Agent)的自然反应,具体领域如下: 1. 2025年AI创作戏剧剧场表演观众反应 2. 2025年人工智能演员机器人舞台表演 3. 2025年AI生成戏剧剧本与人类创造力 4. 2025年AI能否替代人类演员剧场表演辩论 5. 2025年AI剧场表演:真实还是虚假的正统艺术? 该语料库是戏剧AI智能体接受指数(Theatrical AI Agency Index, TAAI)的核心实证数据集——这是首个用于量化剧场观众是否认可AI作为现场表演中合法创造性智能体的计算指标,可为聚焦从业者的AI戏剧研究补充观众接受度数据。 TAAI的计算公式为: TAAI = 戏剧AI智能体接受度(Theatrical Agency Acceptance, TAA) / (TAA + 智能体拒绝度(Agency Refusal, AR)) 其中: 1.0 = 完全的智能体接受度 0.5 = 争议区间 0.0 = 完全的智能体拒绝度 以下为三类基于表演理论的信号类型: 1. 戏剧AI智能体接受度(TAA):作为TAAI的分子项,典型表述包括“AI可以创作戏剧”“AI作为共同创作者”“令人惊艳的AI表演”“AI带来了新的可能性”。理论依据:巴拉德(Barad, 2007)的后人类智能体理论、费舍尔-利希特(Fischer-Lichte, 2008)的变革性表演理论。 2. 智能体拒绝度(AR):作为TAAI的分母配对项,典型表述包括“AI没有灵魂”“这不是真正的戏剧”“无法替代人类演员”“虚假的艺术”。理论依据:本雅明(Benjamin, 1935)的光晕(aura)理论、亚里士多德《诗学》中的模仿说、奥斯兰德(Auslander, 2008)的现场性理论。 3. 本体论焦虑(Ontological Anxiety, OA):属于新型修饰性信号,典型表述包括“这对戏剧界意味着什么?”“谁是创作者?”“这还算艺术吗?”这类哲学层面的不确定性。理论依据:桑塔格(Sontag, 1966)、本雅明(1935)的光晕消解理论;该信号与ASAI、AEDI类的恐惧信号存在明确区别。 主要研究发现: - 总语料规模:共2816条评论,对应约50个视频 - 已完成分类(已计算TAAI值)的样本共10条,占比0.4% - 平均TAAI值为0.6000——智能体接受度略占主导 - 智能体接受度集群样本共6条,占已分类样本的60% - 智能体拒绝度集群样本共4条,占已分类样本的40% - 本体论焦虑信号样本共1条——新型信号得到验证 - 智能体接受度信号密度:0.0129/100 Token(Token) - 智能体拒绝度信号密度:0.0089/100 Token(Token) - 本体论焦虑信号密度:0.0002/100 Token(Token)——罕见但确实存在 领域层面的TAAI结果: - “AI剧场表演:真实还是虚假的正统艺术?”领域:TAAI=0.7500(智能体接受度占主导) - “AI创作戏剧 / 观众反应”领域:TAAI=0.5000 - “AI生成戏剧 / 人类创造力”领域:TAAI=0.5000 关键研究发现: - 点赞数最高的智能体接受度评论(获2713赞):“为什么我感觉AI演员从未有过眼神交流”——该评论结合了智能体接受与诡异谷感知,印证了费舍尔-利希特(2008)的共同在场理论。 - 点赞数最高的智能体拒绝度评论(获1842赞):“有人说我们应该把AI称为‘屁艺术家’,因为这是虚假的艺术”——集体性的反讽式智能体拒绝,印证了本雅明(1935)的光晕理论在自然讨论中的适用性。 - 本体论焦虑信号得到验证:“作为一名以重新演绎《伟大的美国歌本》为职业的从业者,若AI继续[...]”(获214赞)——涉及创作者身份的哲学焦虑,这是首个在剧场观众讨论中被计算验证的本体论焦虑信号。 相关文件: - taai_videos.csv:约50个独立视频的元数据 - taai_comments.csv:2816条原始评论 - taai_results.csv:TAAI评分及信号标注结果 研究方法:YouTube Data API v3。Python 3.12,2026年4月。

提供机构:
Zenodo
创建时间:
2026-04-26
二维码
社区交流群
二维码
科研交流群
商业服务