Musk's AI-Made Odyssey Episode Draws 7.3M Views and Brutal Reviews

Elon Musk's AI-generated rival to Christopher Nolan's The Odyssey has landed on X as a seven-minute episode — drawn 7.3 million views and some of the harshest reviews of the year. The project, made with London startup Wonder Studios using Musk's Grok Imagine video tool, is the first installment of what Musk promised would be a "full-length movie" version of Homer's poem before the year ends. He is now releasing it in episodes instead.

The release matters beyond one billionaire's grudge. It is the most visible test yet of whether frontier AI video generation can produce watchable long-form narrative at a tiny fraction of Hollywood's cost — and whether audiences will accept the results.

Why did Musk make his own Odyssey?

Musk criticized Nolan's blockbuster in the summer and pledged to make a version "true to the art of Homer." He said he would use Grok Imagine to create a "historically accurate" adaptation. His objections centered on casting choices, including Black actor Lupita Nyong'o as Helen of Troy, and he used the hit film as a vehicle for airing his views on diversity, equity and inclusion.

The Tesla chief promised a full-length film before year-end. The pivot to episodic release quietly scales back that commitment — a schedule change nobody at X has explained.

The first episode covers Odysseus agonizing over whether to join the Greek forces sacking Troy. He sails from Ithaca, his home island, with his men crying out for "vengeance," then makes land and fights. The episode ends after 10 years of conflict, with Odysseus holding — ominously — a wooden figurine of a horse.

What did it cost — and how does that compare?

The production economics are the story's sharpest contrast.

• Wonder Studios built the first episode in less than two weeks with a creative team of three, including US-based director Dev Walker.
• The project is understood to have cost tens of thousands of dollars.
• Nolan's The Odyssey cost $250m (£188m), shot across multiple locations over 91 days with a vast crew.
• Nolan's film has earned $1.8bn at the box office.

That gap — roughly three orders of magnitude — is the number AI video advocates will cite. Whether a seven-minute episode can scale to a feature with coherent characters and continuity remains unproven.

Was the team actually trying to make an 'anti-woke' film?

Despite Musk's framing, the people who made it say no. Xavier Collins, co-founder of Wonder Studios, told the Guardian his team were supporters of Nolan's summer blockbuster. Collins said he was a "huge fan" of the film Musk spent months attacking.

"We were not given any direction to make it anti-woke or anything like that," Collins said. "We were given the opportunity by one of the world's best frontier labs to work on one of the best stories ever told."

He also disclosed the actual brief: "help us understand the technology, where it's strong, where it could use improvement."

That framing recasts the project from ideological statement to research exercise. Musk's public positioning and his studio's stated mandate do not match — and neither party has reconciled the two.

How bad are the reviews?

The reception on X has been blunt. A reviewer in the Telegraph wrote: "You can feel your eyeballs twitch as it unfolds in all its sloppy gruesomeness."

Seven-figure view counts on the Grok Imagine account show distribution; the critical response shows the technology's current ceiling. Musk has not said whether he considers the episode a success by his own "historically accurate" standard, and no metrics beyond view counts have been released.

Collins acknowledged specific shortcomings: the historical accuracy of the boats and the episode's rapid pacing are on the fix list. "We are going to take a look at all of the comments and all of the things we could have done better and see if we can improve it," he said. "The goal is that the next episode we make is radically better than the last."

A promise to be "radically better" is easy to make about a first episode widely panned as sloppy. The next installment, with no announced release date, is the real test.

Who is Wonder Studios — and what comes next?

Wonder Studios is a London-based startup with prior AI film work. It has also collaborated with Google DeepMind to produce an AI-generated video for Lewis Capaldi's track Something in the Heavens.

That DeepMind credit matters: it suggests Musk's frontier lab is not the only major AI player using Wonder to probe generative video's limits. The startup functions as a test harness for the technology, not a traditional film house — and its directors say so plainly.

For the AI industry, the episode is a public benchmark. Grok Imagine produced seven minutes of narrative video for tens of thousands of dollars; reviewers judged the craft poor; the audience showed up anyway. Studios and AI labs will read those three facts together when pricing what generated film is worth today.

What's the takeaway?

The stakes are straightforward. If subsequent episodes improve on boats, pacing and coherence, Musk's project becomes evidence that small teams with frontier models can approach theatrical production values. If they don't, it becomes a data point that view counts and production values are still decoupled — and that Musk's promised full-length film, due before year-end, faces the same wall as every other AI feature attempt. The next episode, whenever it arrives, settles which.