[14:58 Wed,4.February 2026 by Thomas Richter] |
The renowned filmmaker Darren Aronofsky (known for *Pi*, *Requiem for a Dream*, *The Wrestler*, *The Whale*, *Black Swan*, among others) has produced a series of short films to mark the upcoming 250th anniversary of the American Revolution, which has caused a stir—not because of their content, but because they were generated entirely by AI.
;) The visuals in the clips were animated entirely by artists using Google&s DeepMind video AI Veo and other AI tools, and then edited, mixed, and color-corrected in post-production. However, the voices—the most lively element of the clips—are spoken by humans, specifically voice actors organized in the American actors& and performers& union SAG-AFTRA—a success of their  successful strike in November 2023. The screenplay is also—still—written by humans, just as the soundtrack was composed by film composer Jordan Dykstra. Aronofsky is thus the first prominent Hollywood director to rely entirely on AI as the sole production tool. Such a (mini-)series, if produced traditionally, would have cost a lot of money due to the historical setting—with AI, it can naturally be realized or rendered much more cheaply: no elaborate historical sets, no actors, no costumes. Every scene can simply be generated via a prompt. But the price for this is high: the photorealistic images are a hybrid between filmed and rendered, and the faces in particular fall into the well-known Uncanny Valley: the eyes are dead, and the facial expressions seem somehow eerie. The look is partly reminiscent of cutscenes in video games; while the individual shots are dramatically lit, they are also aesthetically uninteresting. Looking at the images reminds one of the training process of the video AI, which is reflected in the deeply superficial good looks of the images. The result of inputting highly rated images is an average look of beautiful images, in which the peculiarities of the individual images have been lost in boring framing. Although the quality of the AI images—thanks to steady, rapid progress—is better than it was recently, some characteristics remain: for example, the shots lasting only a few seconds, typical of current video AIs, which automatically dictate the somewhat breathless pace. Ben Bitonti, President of TIME Studios: "This project offers a glimpse into what thoughtful, creative, and artist-led use of AI can look like—not as a replacement for craftsmanship, but as an expansion of the possibilities that allow storytellers to venture down paths that were previously simply closed off to them." Watching the clips, it is not clear why Aronofsky—one of Hollywood&s most idiosyncratic directors—has stepped into the role of an AI pioneer. Sure, a real production would have been extremely more complex, but the AI product presented now seems extremely uninspired—it is a new approach in terms of production costs and speed, but it does not reveal any new creative dimensions. Perhaps it was simply a test of how cheaply one can produce and what the audience can be made to accept. The comments under the videos on YouTube, however, are full of criticism—it will be interesting to see what professional Hollywood will say about his "experiment." Technically speaking, the next AI film will naturally be better—but will it also be worth watching?  more infos at bei time.comdeutsche Version dieser Seite: Darren Aronofsky generiert per KI eine YouTube-Miniserie über die amerikanische Revolution |
 |
|
|