内容来源：https://techcrunch.com/2025/09/06/why-is-an-amazon-backed-ai-startup-making-orson-welles-fan-fiction/

内容总结：

【科技公司启动争议性AI电影修复计划，经典影片《伟大的安巴逊》缺失片段引关注】
一家名为Fable的AI创业公司近日宣布，将利用人工智能技术复原奥逊·威尔斯于1942年执导的经典影片《伟大的安巴逊》中被删减的43分钟片段。该计划由已花费五年时间 digitally 重建威尔斯原作的电影制作人布莱恩·罗斯主导，旨在通过Fable新开发的AI叙事生成模型，在两年内完成对遗失镜头的"再创作"。

值得注意的是，Fable尚未获得该影片的版权许可，其成果可能仅作为技术演示而无法公开放映。该公司自称"AI界的Netflix"，曾获亚马逊Alexa基金投资，此前还因未经授权利用AI生成《南方公园》同人剧集引发争议。

威尔斯遗产管理方对此提出强烈批评，称其"利用威尔斯的创作天才进行炒作"，且未提前与遗产基金会沟通。尽管管理方表示对AI技术本身持开放态度（曾授权使用威尔斯声音模型用于商业配音），但强调此类AI复原仅是无灵魂的"机械复制"，无法替代原作的创造性思维。

影史学者指出，《伟大的安巴逊》作为威尔斯继《公民凯恩》后的第二部作品，因制片方强行删改追加圆满结局而成为"残缺的杰作"，其艺术价值恰恰源于威尔斯独特的作者印记。即便采用"AI+传统拍摄"的混合制作方式（拟通过当代演员面部替换原演员数字复刻），最终成果仍将是对经典的一次重新诠释，而非真正复原失落的原作。

在技术能否弥补艺术遗憾的争议中，该项目再度引发关于AI创作边界与经典作品保护的重要讨论。

中文翻译：

上周五，初创公司Fable公布了一项雄心勃勃却令人费解的计划——用AI技术复原奥逊·威尔斯经典电影《伟大的安巴逊》失传的43分钟片段。这家自称"AI界网飞"、近期获得亚马逊Alexa基金投资的企业，为何要重制这部1942年问世的老电影？

原来该公司开发了一个通过AI提示词生成自制动画的平台。Fable最初使用自有知识产权内容进行训练，但其野心是获得好莱坞IP的改编权限。事实上该平台已被用于制作未经授权的《南方公园》剧集。

如今Fable推出新型AI模型，声称能生成长篇复杂叙事。未来两年内，已花费五年时间数字化重建威尔斯原始愿景的电影人布莱恩·罗斯，将运用该模型重制《伟大的安巴逊》遗失片段。值得注意的是，Fable并未获得该片版权，这意味着这项技术演示很可能永远无法公映。

选择《安巴逊》的原因？若非威尔斯影迷，这个选择确实显得冷门。即便在经典电影爱好者中，威尔斯的第二部作品也始终活在《公民凯恩》的阴影下。前者被誉为影史最佳，后者则因被片方强行删改、添加牵强团圆结局而成为影坛著名的"遗失的杰作"。

这种"遗失的遗憾"正是吸引Fable和罗斯之处。但必须强调：如今我们关注《安巴逊》，完全源于威尔斯——这部电影如何颠覆他的好莱坞生涯，即便残缺版本仍闪耀着其导演才华的光辉。

令人震惊的是，Fable显然未与威尔斯遗产基金会接洽。代表威尔斯女儿比阿特丽斯管理遗产的大卫·里德向《综艺》表示，该项目"企图利用威尔斯的创作天才进行炒作"，最终成果将不过是"缺乏威尔斯式创新思维的机械仿制品"。

里德的批评焦点与其说是重制行为，不如说对方"甚至未履行基本告知义务"。他特别指出："基金会其实已接纳AI技术，正在开发用于商业配音的威尔斯声音模型。"

但我持更保守态度。即便获得威尔斯后人授权，我对新版《安巴逊》仍毫无兴趣，就像厌恶用数字模拟的威尔斯传奇嗓音兜售商品的做法。

威尔斯影迷都知道，这并非首次有人试图替逝者完成遗作。但过往尝试至少使用威尔斯亲自拍摄的素材，而Fable宣称将采用AI与传统制片结合的方案——让当代演员重拍部分场景，再通过数字换脸技术还原原版演员容貌。

尽管在未获版权和遗属祝福的情况下宣布项目十分荒诞，但罗斯至少怀揣致敬威尔斯艺术愿景的真诚。他在声明中痛心疾首地提到："那个失去的四分钟长镜头是影史悲剧"，现存重剪版本仅保留50秒。

我同样为此感到遗憾——但更坚信这是AI无法弥补的永恒损失。无论Fable和罗斯如何完美拼接出移动长镜头，那终将是他们的版本，而非威尔斯的创作；那是约瑟夫·科顿与阿格妮丝·穆尔黑的数字傀儡，而非演员本尊。他们的成品永远不会是八十多年前雷电华公司摧毁的那个原始版本——除非奇迹发生让胶片失而复得，否则真正的《安巴逊》早已永逝。

英文来源：

On Friday, a startup called Fable announced an ambitious, if head-scratching, plan to recreate the lost 43 minutes of Orson Welles’ classic film “The Magnificent Ambersons.”
Why is a startup that bills itself as the “Netflix of AI,” and that recently raised money from Amazon’s Alexa Fund, talking about remaking a movie that was first released in 1942?
Well, the company has built a platform that allows users to create their own cartoons with AI prompts — Fable is starting out with its own intellectual property, but it has ambitions to offer similar capabilities with Hollywood IP. In fact, it’s already been used to create unauthorized “South Park” episodes.
Now Fable is launching a new AI model that can supposedly generate long, complex narratives. Over the next two years, filmmaker Brian Rose — who has already spent five years working to digitally reconstruct Welles’ original vision — plans to use that model to remake the lost footage from “The Magnificent Ambersons.”
Remarkably, Fable has not obtained the rights to the film, making this a prospective tech demo that will probably never be released to the general public.
Why “Ambersons”? If you’re not a Welles-loving cinephile, I’m guessing it sounds like an obscure choice for digital resurrection.
Even among classic movie buffs, Welles’ second film is overshadowed by its older, more famous sibling. While “Citizen Kane” is often called the greatest movie ever made, “Ambersons” is remembered as a lost masterpiece that the studio took out of the director’s hands, dramatically cutting it down and adding an unconvincing happy ending.
Tech and VC heavyweights join the Disrupt 2025 agenda
Netflix, ElevenLabs, Wayve, Sequoia Capital, Elad Gil — just a few of the heavy hitters joining the Disrupt 2025 agenda. They’re here to deliver the insights that fuel startup growth and sharpen your edge. Don’t miss the 20th anniversary of TechCrunch Disrupt, and a chance to learn from the top voices in tech — grab your ticket now and save up to $600+ before prices rise.
Tech and VC heavyweights join the Disrupt 2025 agenda
Netflix, ElevenLabs, Wayve, Sequoia Capital — just a few of the heavy hitters joining the Disrupt 2025 agenda. They’re here to deliver the insights that fuel startup growth and sharpen your edge. Don’t miss the 20th anniversary of TechCrunch Disrupt, and a chance to learn from the top voices in tech — grab your ticket now and save up to $675 before prices rise.
The movie’s reputation — the sense of loss and what could have been — is presumably what interested Fable and Rose. But it’s worth emphasizing that the only reason we care about “The Magnificent Ambersons” today is because of Welles — because of how it derailed his Hollywood career, and how even in its diminished form, it still reveals so much of his filmmaking genius.
That makes it even more astonishing that Fable apparently failed to reach out to Welles’ estate. David Reeder, who handles the estate for Welles’ daughter Beatrice, described the project to Variety as an “attempt to generate publicity on the back of Welles’ creative genius” and said that it will amount to nothing more than “a purely mechanical exercise without any of the uniquely innovative thinking [of] a creative force like Welles.”
Despite Reeder’s criticism, he seems less upset by the idea of attempting to recreate “Ambersons” and more by the fact that the estate was not “even given the courtesy of a heads up.” After all, he noted, “the estate has embraced AI technology to create a voice model intended to be used for VO work with brands.”
I’m not so open-minded. Even if Welles’ heirs were being consulted and compensated, I’d have zero interest in this new “Ambersons,” just as I have zero interest in hearing a digital simulacrum of Welles’s legendary voice being used to hawk new products.
Now, Welles fans know this isn’t the first time other filmmakers have tried to posthumously fix or finish his movies. But at least those attempts used footage that Welles had shot himself. Fable, meanwhile, describes its planned approach as a hybrid of AI and traditional filmmaking — apparently some scenes will be reshot with contemporary actors whose faces will be then swapped for digital recreations of the original cast.
Despite the absurdity of announcing a project like this without the film rights or the blessing of Welles’ daughter, at least Rose seems motivated by a genuine desire to honor Welles’ vision. For example, in a statement about why he wants to recreate the film, Rose mourned the destruction of “a four-minute-long, unbroken moving camera shot whose loss is a tragedy,” with only 50 seconds of the shot remaining in the recut film.
I share his sense of loss — but I also believe this is a tragedy that AI cannot undo.
No matter how convincingly Fable and Rose may be able to stitch together their own version of that tracking shot, it will be their shot, not Welles’, filled with Frankensteined replicas of Joseph Cotten and Agnes Moorehead, not the actors themselves. Their final product will not be Welles’ version of “The Magnificent Ambersons” that RKO destroyed more than 80 years ago. Barring a miraculous rediscovery of lost footage, that version is gone forever.