← Back To All Posts
Date:
February 8, 2023

Content Creation During the Rise of AI

Hardly a day goes by without someone talking about Artificial Intelligence (AI), and apps like ChatGPT , DALL-E , Flawless , Midjourney , Google Bard , or other generative AI technologies are severe threats to original film and TV content creation. Every major news outlet has done stories on AI's potential detriment to the industry.

Pundits have written extensively on the topic, stoking fears that the beginning of the end of the market for human creativity is upon us. But AI does have its advantages.

Impressive Results

The fear is that films could be made without actors, sets, scripts, directors, locations, costumes, and musicians. Yet today, anyone can produce " emotional soundtracks " influenced by any song they upload or any mood or genre they choose by opening a computer program. The public can download phone apps that make any photo look like a famous painter created it. AI-generated special effects have appeared in film, TV, and video games for years , such as in the Marvel Cinematic Universe (MCU).

Alejandro Jodorowsky is an avant-garde filmmaker who incorporated very futuristic costumes and sets in his films. He famously wanted to make a feature adaptation of "Dune" but never did . An article in the New York Times titled " This Film Does Not Exist " describes how the Midjourney AI platform was used to create realistic-looking "production stills" from the fictitious film " Jodorowsky's Tron ." The article discusses how AI was used to interpret the Jodorowsky style and imagine it in a Tron film and how close Canadian Director Johnny Darrell's images came to being something Jodorowsky could have created himself. The result is impressive, but it also highlights AI's primary weakness.

No Copy Is Original

The word "create" should be loosely used when referring to AI. None of the tool's "creations" are genuinely original. Why? AI only knows what it has been taught or shown. By "taught," we mean what examples or data were used to train it. DALL-E can create images "in the style of Monet," and Midjourney can create costumes in the style of Jodorowsky, but neither can create anything new or unquestionably original. AI copies. It mimics. Could AI conceive an original story like "Avatar"? Could it envisage a painting style like Van Gogh or Picasso? Probably not.

To test this hypothesis, we challenged ChatGPT to "Make something up." Its output:

Once upon a time, there was a magical land called Zogonia, where every object had a will of its own and talked to each other. The king of Zogonia was a wise old oak tree named King Oakley, and all the subjects respected and loved him. One day, a wicked sorcerer cast a spell on King Oakley, turning him into a tiny sapling. A brave young adventurer named Lily set out on a journey to find the sorcerer and break the spell. She met many talking objects along the way, each offering their assistance in their own unique way. Finally, after many trials and tribulations, Lily found the sorcerer's castle, defeated him in a duel, and broke the spell, restoring King Oakley to his former glory. The kingdom of Zogonia rejoiced, and Lily was hailed as a hero. From that day forward, she became known as the protector of Zogonia and its magical talking objects.

With no direction whatsoever, ChatGPT produced a children's story about a mythical place with mythical characters, a protagonist, an antagonist, a heroine, and a happy ending. We then asked it to do it again five more times. Each of its responses was not much different from the first. Each followed the same formula of producing characters, plot, and either happy or open-ended outcomes.

The stories were familiar, more of the same outlines, and that's the point. Humans must remain not only in the loop but in the lead. Without human input, AI is incapable of creating original, finished content.

AI can contribute to the creative process by taking ideas generated by human imagination and constructing images or text using that input. It can suggest how to fill storyline gaps or iterative costume design changes, but that's a long way from story concept to finished film. This is especially true when titles travel across borders and cultures.

AI will not take the jobs of writers, directors, or other creatives anytime soon. It will enable those willing to evolve their processes to explore or expand their creative horizons. Humanity allows us to see variations, imagine alternatives, make mistakes, understand perspectives, and recognize the influence of culture and precedence in content creation. AI has none of those capabilities.

What it does have is what director Frank Pavich mentions in his New York Times piece: access to millions of digitized images from across human civilization that directors, concept artists, and film students can use to influence their imaginations and stories. How will that play out?

As Pavich says, "we're about to find out."

Related Insights

YouTube Thumbnails Can Get You in Trouble

Here’s Why Creators Should Pay Attention

When we talk about content compliance on YouTube, most people think of the video content itself — what’s said, what’s shown, and how it’s edited. But there’s another part of the video that carries serious consequences if it violates YouTube policy: the thumbnail.

Thumbnails aren’t just visual hooks — they’re promos and they’re subject to the same content policies as videos. According to YouTube’s official guidelines, thumbnails that contain nudity, sexual content, violent imagery, misleading visuals, or vulgar language can be removed, age-restricted, or lead to a strike on your channel. Repeat offenses can even result in demonetization or channel termination. That’s a steep price to pay for what some may think of as a simple promotional image.

The Hidden Risk in a Single Frame

The challenge? The thumbnail is often selected from the video itself — either manually or auto-generated from a frame. Creators under tight deadlines or managing high-volume channels may not take the time to double-check every frame. They may let the platform choose it automatically. This is where things get risky.

A few seconds of unblurred nudity, a fleeting violent scene, or a misleading expression of shock might seem harmless in motion. But when captured as a still image, those same moments can trigger YouTube’s moderation systems — or worse, violate the platform’s Community Guidelines.

Let’s say your video includes a horror scene with simulated gore. It might pass YouTube’s rules with an age restriction. But if the thumbnail zooms in on a blood-splattered face, that thumbnail could be removed, and your channel could be penalized. Even thumbnails that are simply “too suggestive” or “misleading” can get flagged.

Misleading Thumbnails: Not Just Clickbait — a Violation

Another common mistake is using a thumbnail that implies something the video doesn’t deliver — for example, suggesting nudity, shocking violence, or sexually explicit content that never appears in the video. These aren’t just bad for audience trust; they’re a clear violation of YouTube’s thumbnail policy.

Even if your content is compliant, the wrong thumbnail can cause very real problems.

The Reality for Content Creators

It’s essential to recognize that YouTube’s thumbnail policy doesn’t exist in isolation. It intersects with other rules around child safety, nudity, vulgar language, violence, and more. A thumbnail with vulgar text, even if the video is educational or satirical, may still result in age restrictions or removal. A still frame with a suggestive pose, even if brief and unintended in the video itself, can be enough to get flagged.

And for creators monetizing their work, especially across multiple markets, the risk goes beyond visibility. A flagged thumbnail can reduce ad eligibility, limit reach, or cut off monetization entirely. Worse, a pattern of violations can threaten a channel’s long-term viability.

What’s a Creator to Do?

First, you need to know how to spot the problem and then know what to do about it. Second, you need to know if the changes you make might affect its acceptance in other markets or countries. Only then can you manually scrub through your video looking for risky frames. You can review policies and try to stay up to date on the nuances of what YouTube considers “gratifying” versus “educational” or “documentary.” But doing this at scale — especially for a growing content library — is overwhelming.  

That’s where a tool like SpherexAI can help.

A Smarter Way to Stay Compliant

SpherexAI uses frame-level and scene-level analysis to flag potential compliance issues — not just in your video, but in any frame that could be selected as a thumbnail. Using its patented knowledge graph, which includes every published regulatory and platform rule, it will prepare detailed and accurate edit decision lists that tell you not only what the problem is, but also for each of your target audiences. Whether you're publishing to a single audience or distributing globally, SpherexAI checks your content against YouTube’s policies and localized cultural standards.

For creators trying to grow their brand, monetize their work, and stay in good standing with platforms, that kind of precision can mean the difference between success and a takedown notice.

Want to know if your content is at risk? Learn how SpherexAI can help you protect your channel and optimize every frame — including the thumbnail. Contact us to learn more.

Read Now

Automating Peace of Mind: Navigating YouTube's Global Guidelines with SpherexAI

For media companies distributing content across YouTube, compliance is no longer just a legal requirement—it’s a prerequisite for discoverability, monetization, and channel survival. YouTube enforces strict policies governing child safety, vulgarity, graphic content, and cultural sensitivity. For content owners, ensuring compliance across multiple categories and geographies is a complex and labor-intensive process. To address this issue, SpherexAI provides a scalable solution tailored for any content creator or owner.

YouTube’s Expanding Compliance Landscape

YouTube’s Community Guidelines cover a wide array of regulated categories. Content can be removed or age-restricted—and creators may face penalties—if videos violate policies on:

  • Nudity and sexual content: Content that includes sexually gratifying imagery or non-consensual sexualization is prohibited.
  • Violence and graphic imagery: Footage showing serious injury, bodily fluids, or torture intended to shock viewers can be flagged or removed.
  • Child safety: Content that exploits minors, includes inappropriate family content, or features children in dangerous stunts is not allowed.
  • Illegal or regulated goods: YouTube restricts promotion of firearms, narcotics, and gambling services, among others.

Managing compliance with each of these categories—especially when content is global and multilingual—is a logistical challenge for distributors.

Enter SpherexAI: Precision Compliance Automation at Scale

SpherexAI applies multimodal AI to analyze video content across dialogue, visuals, audio, and metadata. It detects compliance issues not only by scanning for policy violations but also by identifying subtle cultural or regional sensitivities that could result in content removal or limited distribution.

For example, the platform flags:

  • Dialogue with excessive profanity or sexual references, aligned with YouTube’s vulgar language policy.
  • Visuals showing partial nudity, firearm use, or dangerous stunts, which may trigger strikes or age restrictions.
  • Culturally sensitive depictions—such as religious imagery or portrayals of death—that may violate local norms and platform rules.

SpherexAI outputs include timestamped alerts and severity levels, allowing content owners to make targeted edits rather than performing full manual reviews.

Equal Rules for All Creators

Whether you’re a major studio releasing film clips or a digital-first creator uploading your first series, YouTube holds all content publishers to the same standards. Community Guidelines are enforced platform-wide, regardless of a channel’s size, history, or market familiarity.

This presents a significant challenge for new entrants. Many first-time creators or distributors may be unaware that a thumbnail featuring misleading imagery, a prank involving minors, or a scene with unedited drug references can lead to demonetization or a channel strike. But YouTube’s enforcement is uniform: content that violates policy is subject to the same sanctions across the board.

SpherexAI helps level the playing field by equipping every content team—regardless of experience—with access to the same tools used by top studios. Its patented knowledge graph, built on over a decade of regulatory insight and expert human annotation, powers its AI models with unmatched precision. The result: faster reviews, greater accuracy, and fewer costly mistakes.

Cross-Platform, Region-Aware, and Regulation-Ready

Unlike tools focused on metadata or age ratings alone, SpherexAI delivers:

  • Granular analysis: Scene-by-scene breakdowns for violence, vulgarity, sexual content, and self-harm risks.
  • Cultural intelligence: Predictive models assess content suitability across 240+ territories using Spherex’s proprietary “cultural distance” framework.
  • Workflow integration: The platform’s API allows integration into existing supply chains and CMS platforms for automated review at scale.

Reducing Risk, Unlocking Revenue

YouTube’s monetization eligibility hinges on content safety. Channels can be demonetized or de-prioritized in search and recommendation if flagged for repeated violations. Well-known creators Logan Paul, ScreenCulture, and LH Studios have all been sanctioned for violations. By proactively identifying and resolving compliance issues before publishing, SpherexAI empowers content owners to:

  • Avoid strikes or takedowns
  • Retain monetization rights
  • Accelerate time-to-market
  • Protect brand reputation

Conclusion

YouTube is a dynamic platform for global content distribution that requires rigorous adherence to evolving content standards. For studios, broadcasters, and new creators alike, SpherexAI offers an AI-powered safety net automating policy compliance while preserving creative integrity. When SpherexAI is integrated into your production workflow, you can publish confidently at scale, with full compliance, and with no brand risk.

Ready to streamline compliance and expand your YouTube strategy globally?

Book a demo or visit spherex.com to learn how SpherexAI can support your team.

Read Now

Spherex CEO Teresa Phillips Talks Practical AI for Global Content Localization at EnTech Fest

At this year’s DEG EnTech Fest, Spherex CEO and Co-Founder Teresa Phillips joined a panel to explore one of the most practical and impactful uses of AI in entertainment today: localization.

During the session titled “Practical AI For Speed and Savings in Localization,” Phillips shared how Spherex is leveraging AI to deliver “deep video understanding” that accelerates compliance and rating decisions in over 200 markets. As she explained, understanding the context—cultural, visual, and narrative—is crucial in determining whether a piece of content is suitable for audiences worldwide.

“AI can now detect not just what happens in a scene, but how it might be interpreted in different cultural and regulatory environments,” said Phillips. For example, in Scandinavian countries, if a trusted figure, such as a clergy member, commits an unethical act onscreen, it can dramatically impact a film’s age rating. SpherexAI is trained to identify these nuanced moments, flagging them for human review when needed.

Phillips also highlighted the role of AI in augmenting human decision-making, noting that “AI agents can be trained to ask humans the right questions—like whether the drinking in a scene is casual or excessive—ensuring more consistent, scalable evaluations.”

The conversation also acknowledged the broader industry shift that AI is bringing to localization workflows—from quality control (QC) to artwork generation, compliance, and project management. With automation poised to displace some entry-level roles, Phillips raised a key question for the future: “If junior roles are the first to be automated, how do we bring new talent into the industry? We have a responsibility in our organizations to create opportunities for the next generation.”

Joining Phillips on the panel were Silviu Epure (Blu Digital Group), Chris Carey (Iyuno), Kelly Summers (The Sherlock Company), and Duncan Wain (Zoo Digital), offering a 360° view on how AI is transforming the way stories cross borders.

Read Now