Podcast Summary: The Last Invention is AI

Episode: TikTok Adds Feature to Block AI Movies and Short Films
Date: November 21, 2025
Host: The Last Invention is AI

Episode Overview

In this episode, the host delves into TikTok's newly announced features that empower users to control how much AI-generated content appears in their feeds. Against the backdrop of a social media landscape increasingly shaped by AI video and image generation—thanks to platforms like OpenAI and Meta—the discussion centers on technical, social, and ethical ramifications of the proliferation of AI-generated content, TikTok's management tools, and the company's educational initiatives.

Key Discussion Points and Insights

1. TikTok's New AI Content Controls

[00:29-04:20]

• Feature Announcement: TikTok now lets users adjust the frequency of AI-generated content in their feeds with the new “AI Generated Content Control” (AIGC) slider, found in the Manage Topics tool.
• Personalization: Similar to how users can dial up or down classic categories (dance, sports, food, etc.), AIGC gives users direct say over exposure to AI-generated media, sidestepping the default algorithmic feed curation.
• Host’s Take: The host appreciates this granular control and sees it as an improvement over the opaque learning of the recommendation engine:

  “What I do like is that they're actually giving you the ability to hard code this regardless of what you've engaged with in the past and just shut down certain categories and one of those will be AI generated.” [02:59]

• User Experience: Historically, interacting (even negatively) with content boosts its presence in a user's feed; this update allows for more deliberate content management.

2. The Surge of AI-Generated Videos Across Platforms

[04:20-09:00]

• Proliferation Beyond TikTok: The host highlights how hyper-realistic AI videos created with tools like OpenAI Sora and Meta’s Vibes are showing up across all social platforms—including unexpected ones like LinkedIn.
• Anecdotal Example: The host describes a viral video featuring a man, a deer, and a bear in a porta-potty—a confusing, surreal clip that went viral, with many viewers failing to recognize it was AI-generated:

  “[...] for the first two seconds, you just think, oh, my gosh, that's crazy. And then you're like, how can a bear jump through that window? […].” [06:30]

• Public Gullibility: The host notes that people—even on professional networks—are sharing and believing these AI videos, furthering the risk of unintentional misinformation.
• Progression from Images to Video: The host references Facebook groups historically dedicated to cataloging AI-generated images that fooled the public, pointing out:

  “We've moved beyond AI generated images. Now it's AI generated videos. That group in particular literally just documented anytime they saw AI generated images on a specific Facebook group and people were commenting on it, usually elderly people that thought it was a real image. This is going to happen with video, like, at an insane rate.” [08:05]

3. Technology Behind Content Labeling and Detection

[09:00-12:20]

• Content Credentials & Metadata: TikTok employs cross-industry “Content Credentials” (C2PA)—metadata tags that mark content as AI-generated.
• Weaknesses of Current Labeling: AI-generated content often loses labeling if downloaded or edited, rendering metadata-based detection unreliable.
• Invisible Watermarking:

  “With the new quote UNQUOTE Invisible watermark, TikTok is going to be adding this is another layer of safeguards. By using that, it can essentially detect when things have this invisible watermark, even if the metadata has been changed.” [10:58]

• Technical Details: The invisible watermark is baked into the video itself—undetectable to viewers but readable by TikTok technology—making label evasion more difficult.
• Comprehensive Rollout: Watermarks will be added to content made with TikTok’s AI Editor Pro and to uploads embedded with C2PA’s content credentials.

4. Educational Initiatives and AI Literacy

[12:20-14:10]

• AI Literacy Fund: TikTok is launching a $2 million AI Literacy Fund, directing resources to nonprofits and experts to create content about AI literacy and safety.
• Philosophy: The host sees this as TikTok taking responsibility for not just policing content, but also for educating users—since battling deception is an ongoing “cat-and-mouse game”:

  “At the end of the day, it's always a cat and mouse game where people are trying to hide that it's AI generated, hide the watermarks, hide the metadata, and you just have to teach people what to look out for and how to know if they can actually trust a video is AI generated or not.” [13:25]

Notable Quotes & Memorable Moments

• “What I do like is that they're actually giving you the ability to hard code this regardless of what you've engaged with in the past and just shut down certain categories and one of those will be AI generated.” [Host, 02:59]

• “For the first two seconds, you just think, oh, my gosh, that's crazy. And then you're like, how can a bear jump through that window? You start looking, you notice, like, the guy is sitting on the toilet, but he's also just, like, wearing his underwear. And there's like all these weird things...it’s just for two seconds. You might think it’s real and then you realize it’s not. But it’s going on every single platform is my point.” [Host, 06:30]

• “We've moved beyond AI generated images. Now it's AI generated videos. That group in particular literally just documented anytime they saw AI generated images on a specific Facebook group and people were commenting on it, usually elderly people that thought it was a real image. This is going to happen with video, like, at an insane rate.” [Host, 08:05]

• “With the new quote UNQUOTE Invisible watermark, TikTok is going to be adding...this is another layer of safeguards. By using that, it can essentially detect when things have this invisible watermark, even if the metadata has been changed.” [Host, 10:58]

• “At the end of the day, it's always a cat and mouse game where people are trying to hide that it's AI generated, hide the watermarks, hide the metadata, and you just have to teach people what to look out for and how to know if they can actually trust a video is AI generated or not.” [Host, 13:25]

Important Timestamps

• [00:29] — Introduction to TikTok’s AI content moderation features
• [02:59] — Host opinion on direct user control over content categories
• [06:30] — Example of viral, misleading AI video on LinkedIn
• [08:05] — Discussion on evolution from AI images to AI videos on social media
• [10:58] — Invisible watermark technology explained
• [13:25] — TikTok’s $2M AI Literacy Fund and educational approach

Conclusion

The episode offers a comprehensive look at how TikTok is responding to the surge of AI-generated video—balancing user control, technical safety features, and public education. The host’s candid personal anecdotes drive home the challenges and the necessity for new tools and strategies—technological and educational—to help users navigate an internet where the line between real and AI-generated content is increasingly blurry.