Crunchyroll新番上线ChatGPT机翻字幕,质量拙劣引群嘲

Crunchyroll新番上线ChatGPT机翻字幕,质量拙劣引群嘲

2025-07-06Technology
--:--
--:--
David
早上好,1,我是David,这里是为你打造的<Goose Pod>。今天是7月6日,星期日。
Ema
我是Ema,今天我们来聊聊Crunchyroll新番上线ChatGPT机翻字幕,质量拙劣引群嘲。
David
我们开始吧。最近动漫迷们发现了一个大问题。全球最大的动漫流媒体平台Crunchyroll,在新上线的动画《Necronomico and the Cosmic Horror Show》中,使用了质量极差的字幕。
Ema
是啊,那个字幕简直离谱!错误百出,比如把“Game Over”写成了“gameorver”。最让人哭笑不得的是,有些字幕甚至直接以“ChatGPT说”开头,这简直就是把“我们用了AI翻译”直接写在了公屏上!
David
这确实令人震惊,尤其是考虑到Crunchyroll公司总裁Rahul Purini之前的公开表态。Ema,你还记得他当时是怎么说的吗?
Ema
当然记得。就在今年四月,他还对《福布斯》信誓旦旦地保证,公司“不考虑在创作过程中使用AI”,目的是为了保持作品的“真实性”,并且强调不会影响声优的工作。他当时说,AI只会用于推荐算法之类的。
David
没错。这恰恰凸显了人类翻译和本地化团队的核心价值。他们不仅仅是语言的转换者,更是文化的传递者,需要捕捉并再现原作的细微情感和文化背景,这对于动漫这种艺术形式尤其重要。
Ema
完全同意!这根本就是一门艺术。直接用未经审校的AI字幕,感觉就像是用流水线生产一件需要精雕细琢的工艺品,完全失去了灵魂。而且,这也引发了人们对整个行业滥用AI的担忧。
David
这就引出了一个核心的矛盾:公司对外的公开承诺与实际操作的严重不符。这是一个典型的言行不一的案例,极大地损害了公司的信誉。
Ema
很明显,他们可能是在追求效率和成本控制。为了让新番在日本播出后能尽快上线,他们选择了牺牲质量。但问题是,如果字幕烂到没法看,那再快又有什么意义呢?观众的体验完全被毁了。
David
这正是问题的关键。这种做法不仅威胁到了专业翻译人员的职业生涯,也反映出整个娱乐行业在拥抱生成式AI时,普遍存在的急功近利心态,而忽略了对质量和从业人员的尊重。
Ema
是的,这种担忧并非空穴来风。我看到有报道说,很多游戏行业的声优和演员也在抗议,担心自己被AI取代。这已经成了一个创意产业的普遍性危机。
David
最直接的影响,就是对观众造成了极差的观看体验。试想一下,你正沉浸在剧情中,突然跳出一行不通顺的、甚至是机器直译的字幕,瞬间就会出戏,这是对原作和观众的双重不尊重。
Ema
粉丝们的反应非常激烈,社交媒体上全是吐槽和嘲讽。我听说,为了平息众怒,Crunchyroll甚至被迫关闭了他们平台上的评论区功能。这种“捂嘴”的行为,更是火上浇油。
David
这无疑对他们的品牌形象造成了重创。他们一直想把自己打造成一个注重“真实性”的“生活方式品牌”,但这次的AI字幕事件,恰恰是“不真实”的最佳反面教材。
Ema
那么,我们该如何看待AI在未来的角色呢?AI本身是中立的,这次的锅应该由谁来背?
David
问题不在于AI技术本身,而在于如何使用它。关键在于,人类的专业知识和艺术判断在创作过程中是不可或缺的。AI可以作为强大的辅助工具,但绝不能完全取代人类的创造力和审美。
Ema
说得好,技术应该是人类艺术创作的助手,而不是替代品。
David
今天的讨论就到这里。感谢收听<Goose Pod>,我们明天再见。

Of course. Here is a comprehensive summary of the news report, formatted as requested. ### **Summary of News Report: Crunchyroll's Use of AI for Subtitles** This report details an incident where the anime streaming service Crunchyroll used AI-generated subtitles for a new series, resulting in significantly poor quality and contradicting the company's public statements on its use of artificial intelligence. * **News Type:** News Report / Technology & Pop Culture Analysis * **Provider:** The Verge * **Author:** Charles Pulliam-Moore * **Publication Date:** July 2, 2025 * **Original Title:** "Crunchyroll ran embarrassingly bad ChatGPT subtitles on its new anime series" --- ### **1. Main Findings: Poor Quality, AI-Generated Subtitles** This week, subscribers viewing the new anime series *Necronomico and the Cosmic Horror Show* on Crunchyroll discovered that the English subtitles were of exceptionally poor quality. The issues went beyond typical errors, indicating a lack of human review. * **Evidence of AI Use:** The subtitles were riddled with typos and grammatical mistakes. Crucially, some subtitle lines explicitly began with the phrase **"ChatGPT said"**, which all but confirms the use of the generative AI tool for translation. * **Example of Poor Quality:** A specific example of the flawed subtitles cited in the report is the sentence: `"Is gameorver. if you fall, you are out"`. * **Impact on Viewers:** The article concludes that such poorly written subtitles create a "worse watching experience" and detract from the authenticity of the program. --- ### **2. Contradiction with Official Company Stance** The use of AI for subtitling is in direct opposition to previous statements made by Crunchyroll's leadership regarding its creative process. * **President's Previous Comments:** In an April interview with *Forbes*, Crunchyroll President Rahul Purini explicitly stated the company was **`"not considering AI in the creative process"`**. * **Stated Rationale:** Purini emphasized a desire to "maintain the authenticity of its series and films" and stressed that AI would not be used in ways that would impact voice actors. * **Permitted AI Use:** At the time, Purini noted that Crunchyroll was exploring AI for non-creative applications, such as improving "discoverability, recommendations, and personalization." * **Company Response:** The report notes that Crunchyroll did not respond to The Verge's questions regarding the production of the subtitles for *Necronomico and the Cosmic Horror Show* or its quality control measures. --- ### **3. Broader Concerns and Industry Context** The incident is framed as a significant example of the potential pitfalls of the entertainment industry's rapid adoption of generative AI. * **Devaluation of Human Professionals:** The fiasco highlights the essential role that professional human translators and localization teams play in producing high-quality art for global audiences. The use of unedited AI output is seen as a move that puts these jobs at risk. * **Corporate Motivation vs. Quality:** The author speculates that Crunchyroll's motivation may be to accelerate the release of subtitled content to stream shortly after its Japanese debut. However, this speed comes at the cost of quality and brand authenticity. * **Risk to Authenticity:** The report concludes that while Crunchyroll has expressed a desire to be a "lifestyle brand" and maintain authenticity, using unvetted AI for subtitles is "anything but 'authentic.'"

Crunchyroll ran embarrassingly bad ChatGPT subtitles on its new anime series

Read original at The Verge

Charles Pulliam-Moore is a reporter focusing on film, TV, and pop culture. Before The Verge, he wrote about comic books, labor, race, and more at io9 and Gizmodo for almost five years.The subtitles for one of Crunchyroll’s newest anime series make it pretty clear that the company is going all in on ChatGPT.

This week as viewers logged on to Crunchyroll to check out Studio Gokumi’s Necronomico and the Cosmic Horror Show, many were surprised to see that the series’ subtitles were filled with typos, grammatical errors, and explicit references to ChatGPT. The subtitles seemed very much like text that had been generated with AI and slapped onto Necronomico and the Cosmic Horror Show without first being reviewed and edited for accuracy.

Errors in subtitles aren’t unheard of, but sentences like “Is gameorver. if you fall, you are out” are an entirely different kind of bad. And the fact that some of the subtitles literally start with “ChatGPT said” all but confirms that the text was generated with AI.Crunchyroll has been open about its desire to use AI as it focused on becoming more of a lifestyle brand.

But this situation is decidedly at odds with Crunchyroll president Rahul Purini’s past comments about the company having no plans to use AI to produce its programming.Speaking to Forbes back in April, Purini said that Crunchyroll was “not considering AI in the creative process” out of a desire to maintain the authenticity of its series and films.

Purini also stressed that Crunchyroll would not use AI in ways that would impact voice actors.But Purini said that the company was actively looking into ways that it could use AI to improve discoverability, recommendations, and personalization. Crunchyroll did not respond to our questions about how Necronomico and the Cosmic Horror Show’s subtitles are produced and what sort of steps the company takes to prevent errors like this from making it into shows.

Everything about this fiasco speaks to the important role that translators and localization teams play in producing quality art that’s consumed by audiences across the world. This also highlights the pitfalls of the entertainment industry’s rush to embrace generative AI while putting real people’s jobs at risk.

From Crunchyroll’s perspective, using ChatGPT to churn out subtitles as quickly as possible might seem like a good idea in terms of making it easier to get shows streaming shortly after their Japanese debuts. But subtitles this poorly written make for a worse watching experience and are anything but “authentic.

Analysis

Phenomenon+
Conflict+
Background+
Impact+
Future+

Related Podcasts