Earlier this month, Youtuber Rick Blessed called his creator partner Rhett Shull with a question: did one of his recent videos look a little?
The video, which Blessed had uploaded to YouTube shorts on August 5, was a clip of his interview with Pearl Jam Mike McCready guitarist. The creator, who is also a music and multi -stertumer producer, has built a follow -up of 5.1 million subscribers for their guitar related content. He had published the same video on his Instagram page.
Shull, a guitarist who also makes videos mainly about music, said he noticed that something stood out about the YouTube Shorts version: it seemed that it had improved using generative artificial intelligence. The background aspects seemed stained, giving an effect of “oil paint,” he said. Other details, such as blessed hair, seemed especially sharp.
“I have been making videos for a long time and I am someone who spends a lot of time trying to make their videos look like in a certain way, with lighting and color degree and things,” Schull told NBC News in a telephone interview. “And I know how YouTube’s normal compression looks … but what was happening here is very, very different.”
Looking for answers from other creators, Shull published a video entitled “YouTube is using AI to alter the content (and not tell us)”. In the last 11 days, his video has circulated widely through X, where some people have restarted clips of him and labeled YouTube to ask about the claim. A person even published screenshots side by side of Blessed McCready videos to show subtle changes.
In a publication to X last week, YouTube disputed the accusations that it used generative or “compensation”, when artificial intelligence predicts a high resolution image of a lower resolution that uses a deep learning model, in videos of creators.
“We are executing a selected YouTube shorts experiment that uses traditional automatic learning technology to unlock, undo and improve the clarity in the videos during the processing,” the YouTube link account, directed by the head of the YouTube editorial, Rene Ritchie, wrote in X in response to a question of a user who had seen the speech about the Shull video.
It was YouTube’s first official comment about the concerns, which the creators first raised in Reddit in June. Several people had previously published observations similar to Shull and Blessed in R/YouTube.
“I wondered if it was just me,” wrote a Reddit user.
“My God, I also had this today, I hate this, I don’t want a platform to alter my content,” added another.
Shull and others have said that the problem is not just that YouTube could be using AI to alter its content. Many creators have been experiencing with several AI tools for a while, including those that were launched by the platform last year, to help them improve their videos.
The main problem, according to some creators, is that they are not given the option of choosing not to participate.
“It really does not matter if you are using ‘Traditional Automatic Learning’ or” Genai “, it is still altering the videos without prior notice or consent of the content owners. In my opinion, I see this practice as misleading and malicious,” wrote a Reddit user, who was the first to publish on the subject.
The spectators could also distrust the content of the creators, according to Shull, especially in the increase in the “falsification” of AI, which is when the tools of AI are used to generate or modify content without the knowledge of the spectators.
“If someone sees a content that I have done that it seems that it has been altered with AI, the logical conclusion that this person, in my opinion, would skip, ‘oh, well, Rhett is using ia to make videos or alter videos,” Shull said. “Or that I am somehow using it like A, a shortcut or a code of tricks, or that is not real, or that has been false deep. Save many questions.”
It is not the first time that a video giant has fired for allegedly using AI to improve the content.
In January, the spectators made fun of Amazon for using what seemed to be in a poster of the 1922 “Nosferatu” movie. The company did not publicly comment on the reaction.
In February, Netflix also caused controversy with its “HD remassers” of “The Cosby Show” and “a different world”, after the spectators said they noticed facial characteristics deformed in the actors and distorted history. Netflix did not issue a comment on whether he used AI to improve programs.
When more comments were asked about the frustration of some creators, a YouTube spokesman sent NBC News to YouTube Liaison X Post.
In the second part of the Publication X, Ritchie wrote: “YouTube is always working on ways of providing the best possible quality and experience of video, and will continue to take into account the comments of the creators and the spectators as we items and improve in these characteristics.”
Later, in response to a different X user, Ritchie elaborated even more.
“Genai generally refers to technologies such as transformers and large language models, which are relatively new,” he wrote.
“The increase in the scale usually refers to taking a resolution (such as SD/480p) and make it look good at a higher resolution (such as HD/1080p),” the post continued. “This is not using Genai or increasing the scale. It is using the type of automatic learning you experience with computational photography in smartphones, for example, and is not changing the resolution.”
Even so, when Shull’s video picked up more traction, many people in X continued to express their concerns.
“Awesome, now let me turn it off because in reality my shorts is getting worse,” wrote the user of X Captainasthro, who goes through the same username on YouTube, where he publishes about the Star Citizen video game, in response to the Publication X of Ritchie.
“The problem is not what technology is being used,” said Ari Cohn, a lawyer of the first amendment and defamation who serves as the main advisor of technological policy at the Foundation for Individual Rights and Expression (FIRE). “It is that you are changing the content without the permission or even the knowledge of its creator.”
“YouTube has confirmed that it is testing AI in search of clarity in the short ones, but the lack of choice for the creators is causing conversations about trust and authenticity in digital content,” said the strategist of AI and former IP lawyer Wes Henderson in a publication about X.
Shull said he plans to continue uploading videos as he does normally. But, after such a massive response to his video about the improvement of YouTube in some shorts of creators, he said he can start talking more about AI in its content.
“I don’t like it, try to withdraw or be anti-Ai, but rather to have the conversation and say: ‘Hey, what are we doing here? Is it something good? Is it something bad?'” He said. “Because it seems that we are moving very fast and not taking into account many really important problems. Mainly, the main one among them is how this will affect people and their livelihoods in a multitude of different industries, not just mine.”
