Free tools Windows power users keep installed
One-click scans. No signup required.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Yes, it can matter—even when viewers enjoy the content. The issue is not simply whether an AI-made character can entertain people; it is whether audiences understand what they are watching and who made the creative choices behind it.
What Olivia Moore’s TikTok experiment showed
In an August 18, 2026 article for Andreessen Horowitz, Olivia Moore described creating Janie, a fictional 19-year-old prospective sorority member, for TikTok during Alabama’s Bama Rush week. Moore used an image generated with ChatGPT, Minimax 3 for speaking clips, Grok Imagine 1.5 for dancing and soundless clips, and ElevenLabs for sound effects. She says she did not mark the posts as AI-generated, to see whether TikTok would label them itself.
Moore reports that Janie’s first video received almost 100,000 views and that TikTok labeled eight of the 20 videos “Contains AI-generated media.” In her account, the labels had no visible effect on performance or comments. She also describes a separate warning about a fictional “Skirt Shop” sponsorship: TikTok objected to the undisclosed ad, she says, rather than identifying the videos as AI. These are Moore’s reported observations, not independently verified platform analytics or enforcement records.
How viewers reacted when they suspected Janie was AI
Moore says viewers spotted signs such as dresser drawers changing between shots and illegible lettering on a shirt. Some criticized the character; others followed the unfolding story and offered creative feedback after accepting that Janie was artificial. Moore’s impression was that engagement increased once the fiction became apparent. That is a creator’s qualitative account, not evidence that disclosure—or discovery—preserves engagement for every audience or format.
#1 Best Overall
She compares Janie with Tinsley, an AI character she made the previous year. Moore says Tinsley looked more visibly artificial and had a face and voice that varied between clips, while Janie appeared more consistent. Moore reports that nearly 90,000 viewers watched the prior year’s reveal that Tinsley was AI. The comparison is between her own projects, not a controlled test of character realism or audience response.
Why disclosure can change the meaning of a piece
Moore’s argument is that enjoyment is only part of the relationship between creator and audience. People also care about the thought, effort, taste, and care they believe went into a work. If a creator hides AI use, viewers may feel that they were encouraged to mistake a constructed character or process for something else. Moore puts it this way: “When creators hide the tools, people experience AI as a trick.”
Disclosure can supply context rather than erase the creator’s contribution. A person may develop the idea, choose the character’s behavior, guide the tools, and decide what to publish. Moore argues that making those choices visible helps audiences see a human creative role in AI-assisted work: “A person had an idea, made choices and used AI to bring it to life.”
Her analogy is a reproduction of a painting: it can be worth buying, provided the buyer is not led to expect the original. The distinction is not that AI-assisted work has no value. It is that the audience should be able to judge it with an honest understanding of what it is.
What the experiment does—and does not—establish
- It illustrates one creator’s experience. Moore says Janie attracted views, received platform labels on some posts, and continued to draw audience participation. The article does not provide independent verification or a representative audience study.
- It does not show that disclosure never hurts reach. Moore says the labels had no visible effect in this case, but does not compare otherwise identical disclosed and undisclosed posts.
- It separates AI labeling from advertising disclosure. In Moore’s account, the fictional sponsorship warning concerned an undisclosed ad, a different issue from whether content was AI-generated.
- It does not settle every ethical or policy question. The article is a first-person reflection, not a comprehensive analysis of platform rules, legal obligations, or AI’s effects on creative work.
A useful way to judge an AI-made creator
Rather than asking only whether a person on screen is real, consider what the creator is asking the audience to believe. Is the account clearly fictional or presented as a real person? Is the AI use disclosed where it would affect how a viewer interprets the content? Are sponsorships represented transparently? And, once the process is clear, is the work still interesting, funny, moving, or skillfully made?
Moore’s broader point is that generative tools may let people express ideas without traditional production skills, a large budget, or other advantages. That is her perspective, not a measured conclusion about AI’s overall social effects. Her experiment makes a narrower point: an artificial character can still invite real audience participation, but trust depends in part on whether viewers are given context to understand the fiction.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
