Even More Kawaii than Real-Person-Driven VTubers? Understanding How Viewers Perceive AI-Driven VTubers
Abstract: VTubers, digital personas represented by animated avatars, have gained massive popularity. Traditionally, VTubers are operated and voiced by human controllers known as Nakanohito. The reliance on Nakanohito, however, poses risks due to potential personal controversies and operational disruptions. The emergence of AI-driven VTubers offers a new model free from these human constraints. While AI-driven VTubers present benefits such as continuous operation and reduced scandal risk, they also raise questions about authenticity and audience engagement. Therefore, to gain deeper insights, we conduct a case study, investigating viewer perceptions of Neuro-sama, the most popular AI-driven VTuber with 845k followers on Twitch and 753k followers on YouTube. We analyze 108k Reddit posts and 136k YouTube comments, aiming to better understand viewer motivations, how AI constructs the virtual persona, and perceptions of the AI as Nakanohito. Our findings enhance the understanding of AI-driven VTubers and their impact on digital streaming culture.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Explain it Like I'm 14
What is this paper about?
This paper looks at a new kind of online entertainer called an AI-driven VTuber. VTubers are streamers who use animated characters instead of showing their real faces. Traditionally, a real person controls the avatar and speaks for it. The paper studies what happens when an AI, not a human, runs the character. It focuses on the most popular AI VTuber, Neuro-sama, and asks: What do viewers like about her? How do they see her personality? And what do they think about the humans who build and manage her?
What questions did the researchers ask?
The researchers focused on four simple questions:
- What makes AI VTubers appealing to viewers?
- How do viewers feel about “bad” or odd AI behaviors, like being random, rude, or making mistakes?
- How do viewers see the AI VTuber’s personality over time — is it consistent, and how does it change?
- Do viewers think AI can replace the human behind a VTuber? If not, what worries do they have about the AI’s developer?
How did they study it?
To understand viewer opinions, the team analyzed what fans were already saying online.
- Data they used: 108,000 Reddit posts and 136,000 YouTube comments about Neuro-sama, covering 2023 to 2025.
- How they analyzed it (in everyday terms):
- First, they collected lots of comments from Reddit and YouTube.
- Then, they used a smart AI helper to tag which comments matched each question (like a librarian putting sticky notes on the right pages).
- Next, they used “topic modeling,” which is like a sorting machine that groups comments talking about the same thing into piles.
- Finally, the researchers read examples from each pile and gave them clear, human-friendly labels so the results made sense.
- They also double-checked the AI helper by hand on random samples and found it was accurate about 96% of the time.
What did they find?
1) What attracts people to AI VTubers
- AI + human chemistry is a big draw. Viewers love when Neuro-sama interacts with humans — her developer (Vedal) and other VTubers. Many fans even describe fun “family-like” roles (for example, Vedal as a patient dad to a mischievous AI). This human connection helps the AI feel “alive.”
- AI + AI chemistry is also fun. Neuro-sama has an “evil twin” called Evil Neuro. Their sister-like rivalry and teamwork are entertaining on their own, even without human guests.
- Randomness and unpredictability are features, not bugs. In normal apps, an AI being chaotic is a problem. But for entertainment, fans enjoy Neuro’s surprise, silliness, and weirdness — it’s part of her charm.
- Cuteness still matters. Just like human-run VTubers, “kawaii” looks and behavior are a big reason people watch.
- Synthetic voice can be appealing. Even though her voice is computer-generated, many fans think it sounds cool, unique, and even beautiful when singing — similar to how people enjoy Vocaloid music like Hatsune Miku.
- The community is a huge part of the fun. Fans make art, memes, and clips that keep the excitement going between streams and attract new viewers.
Why this matters: It shows that AI entertainment doesn’t need to act perfectly human to succeed. In fact, “AI-ness” — including its quirks — can be a strength when the goal is fun, not accuracy.
2) How viewers treat “bad” or odd AI behavior
- Quirks become personality. Forgetting things, getting off-topic, or repeating herself — fans often turn these into running jokes or character traits instead of seeing them as mistakes.
- Rudeness is often treated playfully. Neuro-sama sometimes swears or “roasts” people. Viewers usually don’t take offense because they see it as an AI doing bits, not a human being mean. That said, this area always needs careful moderation to avoid real harm.
- Collabs work differently with AI. When humans collaborate with Neuro-sama, fans don’t expect smooth teamwork. The fun is watching humans react to her chaos — trying (and failing) to keep things on track becomes the comedy.
Why this matters: The rules are different when the performer is an AI. Fans reframe glitches as jokes and “attitude” as part of the act — which changes how creators can design AI characters and shows.
3) How viewers see the AI’s personality over time
- The character is co-created with the audience. Neuro-sama’s identity isn’t just programmed in and left alone. Fans build “lore,” create memes, and connect dots across streams, which shapes a stable-but-evolving personality over time. In short, the community helps write the character together with the AI.
Why this matters: Successful AI VTubers aren’t just software; they’re the center of a social world where fans, the developer, and collaborators grow the character together.
4) Can AI replace the human behind the avatar?
- Viewers still see humans as essential. Many fans think the developer (Vedal), human collaborators, and the community are the real backbone of the content. The AI shines most when it’s part of a human-guided show.
- Old worries don’t vanish — they move. Traditional VTubers can be hurt by the real-life issues of the human behind the avatar. With AI VTubers, viewers worry about the developer instead: their health, choices, relationships, and management style can still affect the show’s stability.
Why this matters: AI doesn’t remove humans from the picture. It shifts which humans matter most and how.
What does this mean going forward?
- AI entertainers can thrive by being delightfully non-human. Leaning into unpredictability, playful chaos, and unique voices can make streams more fun — as long as safety checks and moderation are in place.
- Keep a human in the loop. Developers, co-streamers, and fan communities are key to making AI performances feel lively and meaningful.
- Design for collaboration and community. The best AI VTuber shows use human–AI chemistry, recurring relationships, and fan-made content to build a living “universe” that people want to return to.
- Support the people behind the AI. Even without a face on camera, real humans still carry pressure. Taking care of developers and collaborators helps the AI performer succeed long-term.
In short: The study shows that AI VTubers aren’t replacing people — they’re becoming the playful center of a bigger team effort. The magic happens when AI weirdness, human warmth, and fan creativity all work together.
Collections
Sign up for free to add this paper to one or more collections.