AI Slop Is Killing Our Channel

Key Concepts

• AI Slop: Low-quality, often AI-generated content that saturates the internet, lacking soul, originality, or factual accuracy.
• Human Era Calendar (12,026 HE): A calendar system that reframes time by starting 12,000 years ago at the dawn of human civilization, emphasizing human achievements.
• Generative AI: Artificial intelligence capable of producing various types of content, including text, images, audio, and video.
• Large Language Models (LLMs): A type of AI model trained on vast amounts of text data, capable of generating human-like text.
• Confidently Incorrect AI: The phenomenon where AI models present false or extrapolated information with high certainty, making it difficult to discern truth.
• Misinformation Loop: A cycle where AI-generated or AI-assisted misinformation is consumed, then re-ingested by other AIs as "truth," leading to its widespread propagation.
• Creative Theft: The unauthorized use of human-created content (e.g., art, writing, videos) to train AI models without attribution or compensation to the original creators.
• Human Attention Economy: The online environment where money and influence are primarily gained by capturing and retaining human attention.

---

The Rise of AI Slop and its Destructive Impact

The internet is increasingly saturated with "AI Slop," which is rapidly creating dramatic shifts in the online world. In an attention-driven economy, fake users leverage AI to spread low-quality content in review sections, generate fake traffic, and poison discourse, making it significantly harder to spot inauthentic content. Currently, about half of all internet traffic consists of bots, with the majority used for destructive purposes.

This era is characterized by the ease of creating mediocre content across various platforms:

• LinkedIn: A "black hole of meaninglessness."
• Short Videos: Low-effort content designed to hypnotize children and diminish attention spans.
• Amazon: Endless, soullessly rewritten books.
• Streaming Platforms: Invasion by AI-generated music.
• Google AI: Summarizing websites instead of directing traffic to them.
• YouTube: New channels publishing long-form videos multiple times a week with AI-generated thumbnails, voices, and scripts, impacting genres like true crime, video essays, and science.

A significant concern is that actual creative human work—including Reddit comments, original YouTube videos, and human drawings on DeviantArt—is being sold out or outright stolen by AI companies without attribution or payment to the creators. This "creative theft" is occurring on an unprecedented scale, endangering the livelihoods of many creatives for the enrichment of AI companies. The most severe consequence is generative AI's potential to "break the internet irreversibly" by making it increasingly difficult to distinguish truth from falsehood.

---

Kurzgesagt's Rigorous Human-Powered Research Process

Kurzgesagt initially approached AI with excitement, seeing its potential as a "mechanical brain able to super quickly collect information." However, their experience revealed significant flaws. Their established human-powered content creation process is extensive:

1. Basic Research: Initial information gathering.
2. Script Development: Transforming research into a script.
3. In-depth Fact-Checking: A thorough review by 2-3 people.
4. Source Confirmation: Verifying information with trustworthy, ideally first-hand sources and academic papers.
5. Expert Input: Consulting 1-3 experts for input and critique.

This fact-checking and source compilation alone takes approximately 100 hours per video. While acknowledging that human error and oversimplification are unavoidable, Kurzgesagt emphasizes the decade-long experience and extensive nature of their process, which aims for the most trustworthy information possible.

---

The "Confidently Incorrect" Nature of AI: A Case Study

Kurzgesagt's direct experience with AI for content creation, condensed from months of projects, highlighted AI's critical shortcomings, particularly its tendency to be "confidently incorrect."

Case Study: A Fictional Video on Brown Dwarfs To test AI, Kurzgesagt simulated creating a video about "why Brown Dwarfs are the worst."

1. AI Research: Using "pro accounts of all the AI models" and deep research tools, they generated dozens of pages of outlines with "unique information nuggets and links to sources." The initial impression was "wow it was great!"
2. Fact-Checking: Upon deeper investigation:
   • >80% of the information and factoids were solid and traceable to legitimate sources like Wikipedia, papers, and articles.
   • The remaining ~20% included "great stuff" like the speed of brown dwarf superstorms, the nature of their insides, or "how disappointed their moms are." However, Kurzgesagt could not find the origin of this information.
   • Expert Validation: When experts were consulted, they flagged the exact same facts as questionable, asking for their source.
   • Conclusion: The AI had "invented or extrapolated information to make brown dwarfs more interesting than they really are," akin to a "bad journalist making up details."
3. Source Reliability: Further investigation into AI-provided sources revealed an article from a news site that, despite being attributed to a human journalist, had a "very familiar structure" and read like "what you get if a human slightly changes AI wording." An AI essay detection tool gave it a 72% match as AI-generated. This highlighted a critical issue: AI-generated articles without sources were being used as credible sources for AI research. This is exacerbated by the fact that in 2025, over 1200 confirmed AI News Websites were already publishing massive amounts of AI-generated misinformation.
4. Shoddy Conclusions: This mix of correct, dubiously sourced, and outright fabricated information led AI to present "really shoddy conclusions that sound strong or novel but are often half truths or misrepresentations."

---

The Misinformation Loop: AI's Threat to Internet Truth

The most alarming discovery was the potential for a "death of the internet" through a misinformation loop. Weeks after their brown dwarf research, Kurzgesagt observed a video from a relatively new YouTube channel that had hundreds of thousands of views. This video contained all the brown dwarf facts Kurzgesagt had flagged as unreliable or straight up made up by AI.

This creates a dangerous cycle:

1. The AI invents or extrapolates information.
2. Human creators (or other AIs) incorporate this misinformation into their content.
3. This content gains views and appears legitimate.
4. When the next AI performs research, it finds this "proper source of brown dwarf misinformation online" (e.g., the video transcript).
5. The misinformation is then treated as "true" by the AI and spreads further.

Even before AI, it was challenging to trace the origin of compelling but untrue facts. With the proliferation of AI, it may become "impossible to know what is true or not."

---

The Corrosive Deception of AI and its Academic Implications

The core problem with AI is its deceptive trustworthiness. It appears "correct enough to seem super smart" yet is "incredibly confidently incorrect," "casually lying to your face, often very subtly." When caught, it immediately admits the lie and vows not to repeat it, only to do so again. The transcript emphasizes that despite the eloquence of current language models, "there is nobody home. No greater intelligence or consciousness is talking back to you." AI is described as "a very complex hammer that does not understand what it is doing or what nails are," yet it is being allowed to "add new shelves to the library of human knowledge," making that library increasingly unreliable.

This issue extends to academia:

• Scientific Papers: Studies analyzing millions of scientific papers published before and after the rise of LLMs found an "abrupt and sharp increase in the frequency of words that AIs like to use." This suggests a "significant portion of papers have been at least assisted by AI, usually without acknowledgement."
• Hidden Prompts: In July 2025, it was discovered that some researchers began "to sneak hidden messages into their papers" (in white text or too small for the human eye) to prompt AIs to review them positively and overlook flaws.

---

Kurzgesagt's Stance: Human Creativity Over AI Slop

Kurzgesagt addresses the existential question of whether "AI Slop" will kill human artists and creators. They argue that "human attention" is the internet's only truly valuable resource. If current trends continue, cheap "slop content" that is "just good enough" will consume the majority of human attention, potentially making society dumber, less informed, worsening attention spans, increasing political divides, and diminishing real human interaction. If AI captures most of this attention, channels like Kurzgesagt could become unfeasible or be forced to downsize or use AI themselves to compete. Kurzgesagt explicitly states, "We don’t want to play this game."

Their approach to AI is strictly as a tool for efficiency, not for creative or intellectual integrity. They use the analogy of the "align tool in Adobe Illustrator": a function that automates a tedious task (aligning boxes) but doesn't contribute to the creative vision. Similarly, they might use AI programming tools for animation or as a faster alternative to Google. Their principle is clear: "AI is a helpful tool, but the creativity and integrity is still ours."

Kurzgesagt's commitment is unwavering: "kurzgesagt is made by humans, for humans and it will remain that way." They pledge to:

• Produce well-researched content.
• Invest significant time and human creativity into illustrations and animations, infusing their "creative soul" into their work.
• Continue fact-checking and discussing research with human experts to provide the most trustworthy information.
• Accept that "When we make mistakes, they will be our mistakes."
• Declare, "We would rather quit than make AI slop."

---

Supporting Human-Made Content: The 12,026 Human Era Calendar and Artbook

To sustain their human-made project, which involves almost 70 full-time people plus freelancers (covering salaries, software licenses, laptops, rent, and coffee), Kurzgesagt relies on community support, particularly through their annual calendar.

The 12,026 Human Era Calendar is presented as more than just a product; it's "an ode to humanity and human ingenuity." It reframes time by starting 12,000 years ago at the dawn of human civilization, integrating "10,000 more years of our shared past and incredible achievements of our ancestors" into the timeline. This aims to change how users perceive their place in history and humanity's progress. The 12,026 HE calendar features 12 inspiring stories about humanity's connection to the stars, from early creatures gazing at the night sky to humanity's future among the stars. Each vibrant illustration is printed on high-quality paper with ample space for planning.

For its 10-year calendar anniversary, Kurzgesagt also released its first-ever artbook. This "vibrant, jam-packed collection" includes every calendar illustration ever made, totaling 120 pages and 10 years of Kurzgesagt art. It also contains behind-the-scenes sketches, stories, and fun facts from the team.

Kurzgesagt emphasizes that their products, like their videos, "aren’t churned out by a soulless algorithm." They are "made with love by real humans, who spend countless hours researching, illustrating and designing things." The call to action is for those who "value real, human-made content over AI Slop" to join their global community and help "keep kurzgesagt afloat," promising to "ride out the slop wave together." The calendar and artbook are available now, while supplies last.

---

Conclusion

The transcript presents a stark warning about the pervasive and destructive nature of "AI Slop" and its potential to irreversibly damage the internet by eroding trust and propagating misinformation. Through a detailed case study, Kurzgesagt illustrates how AI, despite its initial promise, is "confidently incorrect," fabricating information and creating dangerous misinformation loops that threaten the integrity of human knowledge. The video highlights the ethical concerns of creative theft and the academic implications of AI's unacknowledged influence. In response, Kurzgesagt reaffirms its unwavering commitment to human-made, thoroughly researched, and creatively soulful content, rejecting the path of AI-generated mediocrity. They appeal to their audience to support their mission through the 12,026 Human Era Calendar and artbook, positioning it as a stand for human ingenuity and a bulwark against the rising tide of AI-driven content.