Sora Is Here: AI Video Generation and the Content Creation Shakeup

OpenAI released Sora, their AI video generation model, to a limited group of creators this month. The samples are impressive—60-second video sequences from text prompts, with reasonable coherence and visual quality. The internet is predictably freaking out about AI replacing video producers. As with most AI announcements, the reality is more nuanced and more interesting.

Here's how I see down what Sora actually does, where it's genuinely useful, and where the hype is getting ahead of reality.

What Sora Can Do

Sora takes a text description and generates a 60-second video. Not a slideshow with music. An actual video with camera movements, lighting, and scene composition. The quality varies—some outputs look cinematic, others look like they were rendered in 2015. But the capability is real.

Examples from the demo: "A beautiful silhouette of a man playing saxophone during a golden sunset." "A group of people shopping for fresh produce at a farmer's market." Sora can generate these kinds of scenes reasonably well.

What it struggles with: complex logic, accurate text rendering, consistent object behavior across a sequence, and anything requiring precise spatial reasoning. If you ask it to show a specific number of people doing something in a specific order, it often gets details wrong.

The Hype vs. Reality

Hype: "AI is going to replace video producers."

Reality: Sora is a tool that might reduce the amount of raw production work, similar to how Photoshop reduced the need for darkroom technicians. But creating compelling video—whether AI-generated or human-created—still requires creative direction, storytelling, and taste. Sora generates a video. It doesn't tell a story.

Hype: "You can generate any video you can imagine."

Reality: You can generate videos for specific, well-defined scenarios. Complex scenarios, narrative arcs, or videos that need to be perfectly accurate? You're still going to hire a human filmmaker or use traditional production.

Hype: "This is going to destroy the video industry."

Reality: This is going to change the economics of certain types of video work. Stock footage. B-roll. Quick promotional videos. These might shift from hiring crews to using AI generation. But the high-end, high-stakes video work—commercials, documentaries, films—will still be human-created for years.

What This Means for Professional Services Firms

Here's the honest answer: Sora is probably not immediately useful for most professional services work. Your firm doesn't produce a lot of video content, so the ability to generate video from text doesn't change your business model.

But there are some uses worth thinking about:

Client Communications Could you use Sora to generate short explainer videos showing how a process works? Possibly. "A lawyer explaining the deposition process step-by-step" could be generated, edited, and used in client onboarding materials. Lower production cost than hiring a crew.

Training and Onboarding New associate training videos. Process walkthroughs. Interview preparation coaching videos. These are formats where you don't need Hollywood-level production quality. Sora could work.

Marketing Materials Short promotional videos for your website or LinkedIn. Sora could generate these. The quality bar is lower than you'd expect. Some slickness around the edges might actually give you an edge over competitors with no video at all.

None of these are significant shifts. But they might make video more accessible to firms that currently don't produce video because the cost and complexity is too high.

The Liability and Authenticity Question

Here's something people aren't talking about: if a firm posts AI-generated video pretending it's real footage (of your offices, your team, your actual client work), that's dishonest and potentially actionable. There's implicit deception.

If you use Sora to generate a training video or an illustrative explainer, that's fine. But if you use it to generate footage of your actual practice, your actual lawyers, doing actual work, without disclosing that it's AI-generated—you're crossing a line.

The copyright implications are also unclear. Sora is trained on copyrighted video. If Sora generates a video that looks suspiciously similar to someone else's copyrighted work, who's liable? OpenAI? You? Both? That's going to be litigated extensively.

My Take

Sora is genuinely impressive from a technical standpoint. It's a real advance in generative AI. But it's not yet at the point where it replaces thoughtful video production, and for professional services, the use cases are limited and mostly supplementary.

The interesting question isn't "does Sora replace video producers?" It's "how does the combination of Sora plus human creativity produce better work than either alone?" That's where the value is.

For now, if you're a professional services firm: watch this space, but don't restructure your communications plan around it yet. In 18 months, Sora will be more capable and more integrated into workflows, and the use cases will be clearer.