YouTube: The Hidden Cost of AI Videos, a Creator's Experience

A Professional Turn Towards YouTube

Jonathan made a radical choice by leaving his job in customer service to embark on a bold adventure: creating AI-generated videos on YouTube. This decision, driven by his passion for AI, led him to discover the often-overlooked costs associated with this activity.

The Costs of Producing AI Videos

Producing videos using artificial intelligence is not a cheap endeavor. Jonathan must invest between £300 and £800 per video, which is approximately $400 to $1,070. These costs primarily include expenses related to the repeated generation of AI content, a process that turns out to be more expensive than it initially appears.

From Idea to YouTube Channel

Before fully committing to content creation, Jonathan worked in customer service. He spent his free time experimenting with AI, despite lacking formal training in filmmaking or professional video creation. In June 2025, he decided to leave his job to fully dedicate himself to his passion. He already had a presence on social media with channels focused on history and enthusiastically launched his YouTube channel, Chloe VS History.

The Creative Process Behind the Videos

Jonathan's channel features Chloe, an AI-generated character who explores significant historical events, such as the maiden voyage of the Titanic or the eruption of Vesuvius in Pompeii. Chloe is presented as a modern vlogger, allowing viewers to immerse themselves in these fascinating historical contexts.

To create these videos, Jonathan uses a suite of AI tools. He relies on Claude to develop the idea and structure the script. The images are generated using PAI 2 from Utopai Studios, with the assistance of Nano Banana Pro and ChatGPT 2.0. The images are then transformed into video clips via Seedance 2.0, and an AI voice model ensures the consistency of Chloe's voice.

The Challenges of AI Content Creation

Creating the character of Chloe was relatively quick, requiring between five and ten versions before finding the right one. However, Jonathan spent much more time refining her personality and dialogues to make her appear authentic, like a real influencer. This attention to detail paid off, to the point where many viewers wonder if Chloe is a real person.

Despite the automation, the videos require rigorous creative direction. Jonathan does not settle for the raw results provided by the software; each scene is carefully reviewed, adjusted, and edited to achieve the desired quality, akin to a traditional film. Historical content presents additional challenges, as AI models can introduce anachronisms, such as modern streetlights in a video about ancient Rome.

A Demanding Production Process

Producing a single long video can span several weeks. This complex process includes scriptwriting, image generation, video creation, voice work, numerous revisions, and post-production.

The Costs Associated with Creation

Long videos on Jonathan's channel typically cost between £300 and £800. The final cost depends on the complexity of the subject and the number of revisions needed to achieve a satisfactory result. Unlike traditional YouTube videos, where expenses may include equipment or travel, the majority of Jonathan's costs come from generating and regenerating content through AI models. Each revision incurs a cost, with approximately $3 per 15-second clip, and a video may require 10 to 15 revisions, which drives up the bill.

The Evolution of Activity on Platforms

Initially, Jonathan attempted to monetize his content on Instagram, TikTok, and YouTube, but only YouTube truly worked. Instagram did not generate revenue as the platform does not pay for views, and TikTok was not a major source of income since Jonathan was not accepted into the creator rewards program. Even on YouTube, monetization took time to establish.

It was only by producing longer videos that Jonathan began to see an improvement in his revenue. Long videos require a different approach to capture viewers' attention for 15 or 20 minutes, unlike short videos that simply need to interrupt scrolling. The more viewers watched until the end and engaged with the long content, the more YouTube recommended these videos.

This recommendation cycle had a snowball effect on the channel. As long videos gained popularity, they also drew attention to Jonathan's Shorts, which had not managed to break through on YouTube on their own.

Conclusion: The Key to Success

Jonathan understood that access to AI tools is not enough to distinguish successful creators from others. While anyone can sign up for software and generate images or videos, the real challenge lies in understanding audience expectations and establishing a process that consistently meets them.