The transformation happened so gradually that we didn’t even notice when fan spaces went from creating for community appreciation into optimizing for algorithmic visibility. A shift occurred at some point, replacing the Star Trek fan who wrote fanfiction for fun in 1995 with the BookTok creator who plans their posts for optimal engagement in 2025.

I wanted to investigate fandom, how it is being taken apart, and what we are losing as AI rapidly changes it. This is what I found.

What is Authentic Fandom?

Before algorithmic curation, fan communities operated according to what anthropologists call a “gift economy” – creative labor freely shared for community appreciation rather than metrics.

Nancy K. Baym’s ethnographic study of a Usenet newsgroup that grew from 11 posts in 1984 to 2,412 by 1993 documents exactly what the algorithms we now know killed. This was not passive consumption by any means. Through teamwork in understanding, sharing information, offering critiques, and forming real friendships, members developed a “community of practice,” in Baym’s words.

The participation patterns revealed that just 33% of members posted only once, and a core 2% writing 100-199 messages generated 22% of content. They weren’t influencers focused on numbers, but rather people who spoke for the community, had distinct views, and built bonds with other fans. One engineer pushed back against stereotypes of soap opera fans:

This gift economy operated on shared norms that governed legitimate versus illegitimate cultural production. Fans created terms like “character rape” to criticize interpretations that opposed a character’s known personality, whether from other fans or the show itself. The intense debates over Kirk/Spock slash fiction are a well known example of this negotiation over appropriate reworking of source material.

Theodor Adorno provided theoretical grounding for these practices decades before scholars celebrated them. His ideas about “use-value” versus “exchange-value” explained how fans prioritized personal connection over money.

In a reflection of these principles, the early Star Trek fandom established practices that would define fan culture for generations – female fans consistently rewrote this “masculine” space-opera through “feminine” genres, and prioritized communal engagement amongst the group. They even invented fan conventions as we know them today!

TikTok’s Structural Transformation

TikTok restructured the relationship between fans, content, and community through what researchers deem “the algorithmized self.”

The way we use social media traditionally is based on a “networked self,” where who we are is built through our social circles. You follow friends and communities; your feed reflects those relationships.

TikTok inverts this entirely. Users primarily interact with algorithmic reflections of themselves rather than other people, deriving identity from engagement with curated content divorced from social relationships.

This platform uses different technical tools to examine captions, hashtags, audio, images, viewing duration, user activity, and how quickly people scroll. Research shows the TikTok algorithm identifies user vulnerabilities and interests in under 40 minutes – unprecedented speed compared to other platforms. The result is an “algorithmic closeness” where users relate more intimately to the recommendation system than to other humans.

A Singapore study of 25 TikTok users aged 21-26 found that all participants relied heavily on the For You Page algorithm, with most rarely liking or commenting on content. When they did engage, these features functioned primarily for bookmarking rather than social interaction. One user reported creating three separate accounts before successfully “training” the algorithm to her preferences, which is a significant effort invested in achieving desired algorithmic outcomes rather than connecting with people.

Users find algorithmic curation convenient, and at the same time, they employ strategies to preserve their own control. They accept data collection as “the price of personalization”, but do express anxiety about algorithmic influence. One participant articulated this boundary by saying: “Unless I can really see how TikTok is controlling me, I will just continue to use it.”

The platform’s architecture shapes community formation in ways that diverge from organic development. Analysis of BookTok – TikTok’s reading community with 87.4 billion hashtag views – examined the top 150 videos and found only 3 creators of color and 3 male-presenting creators, despite both researchers and community members knowing the actual community is far more diverse.

Evidently, TikTok doesn’t merely host BookTok. It constructs what BookTok appears to be through algorithmic choices about which content gets amplified to new users. In-video search bars, playlist features, hashtag systems, and account tagging are the mechanisms shaping what these communities become.

What Dies When Gift Economies Collapse

The shift from community-driven to algorithm-driven fandom produces specific casualties that alter fan practice.

The Death of Reciprocity

There’s now a phenomenon that creates social pressure for fan responses perfectly timed to commercial release schedules, pushing aside more collaborative and sustained fan work, and it’s only going to get worse with the increase of AI.

A whopping 62% of Americans are using generative AI already. Whether or not the bubble will burst remains to be seen (though it is highly likely), but regardless – when robots can generate content instantly, the human labor and reciprocity that sustained any and all gift economies is devalued.

Community Fragmentation Despite Mass Participation

The paradox of algorithmic fandom reveals itself through profound isolation despite our unprecedented ability for communication – users are choosing repackaged versions of themselves rather than genuine connections.

A “consensus illusion” provides perfect cover for this algorithmic manipulation of community. The Spanish Love Deception appeared frequently in BookTok searches despite predominantly negative mentions. Algorithmic systems do not distinguish positive from negative attention – they amplify all engagement, creating false impressions of popularity.

Cross-platform sharing patterns show that users frequently share TikTok content through WhatsApp and Telegram because friends cannot see the same content organically due to personalized curation, forcing active sharing across platforms rather than allowing for organic discovery within different communities.

Performance Replacing Expression

The overload of customization options available to TikTok users are creating psychological barriers to authentic participation, with research showing users think “having many customization features makes it quite difficult for me to create, the barrier to entry is quite high.” Too many options demand cognitive resources that discourage the spontaneous creation characterizing earlier fan cultures.

Users report losing track of time, finding it “easy to just get carried away for hours” in ways differing from the intentional participation of pre-algorithmic fandom. TikTok is constantly offering new things to keep you entertained, but not building a sense of community.

Strategic behavior replaces authentic expression. Users employ hashtags for visibility over accuracy, follow creators to “train” algorithms rather than build relationships, and time content for algorithmic favor rather than community conversation.

Platform Blindness and Weaponized Fannishness

Social media companies’ year-end reports present visions that “there’s nothing wrong with social media…it’s all about fostering love and admiration”, all deliberately designed with bright colors and chunky graphics to convey unmitigated positivity.

Reality differs sharply. UC Berkeley’s Abigail De Kosnik noted that platforms “completely failed to address the more questionable trends within fandom,” whether offensive content or political disinformation. Platforms profit from engagement regardless of content type, then selectively report only positive metrics while denouncing responsibility.

BookTok’s demographic homogeneity in algorithmic promotion does have real-world consequences. When platforms reinforce inequalities rather than organic community formation, this of course impacts publishing decisions, book sales, and whose stories get mainstream promotion. Bookstores now feature “popular on BookTok” displays, meaning this algorithmic curation directly shapes which authors get physical shelf space. Those algorithms – to little surprise – overwhelmingly promote white, heterosexual-presenting creators, and therefore systemic publishing inequalities get further reinforced.

Worst of all, fannish practices – interpretation, community coordination, creative productivity – now have been taken and bastardized by conspiracy theorists. Research on “fanspiracies” demonstrates how QAnon, the January 6th insurrection, and toxic political “fandom” all employ fan-like practices now. When an AMC Entertainment CEO accidentally tweeted the letter “M,” theorists spent hours “decoding” hidden meaning, moving stock prices. His clarification that it was a mistake and joking followup tweet “Y” only generated followup theories – identical to how fans analyze cryptic showrunner posts or Taylor Swift’s captions.

AI as Accelerant

Just as algorithmic curation dismantled gift economies, artificial intelligence threatens to complete the transformation by removing human labor entirely.

According to MBO, 74% of independent workers use generative AI, with 45% feeling “hopeful” versus only 23% feeling worried. Users claim average productivity gains equivalent to 9 hours per week saved, using AI to gain insights, automate tasks, and complete labor.

But fandom-specific impacts remain up in the air. AI-generated Harry Potter fanfiction (“botfic”) went viral. An AI Freddie Mercury cover of “Thriller” reached nearly 1.6 million views in 9 months. Teresa Teng has been “recreated” through thousands of AI covers on Bilibili nearly three decades after her death.

However, these things were not met with unwavering acceptance. Fans typically accept virtual performers like Hatsune Miku and VTubers as “nonhuman” but reject AI-mediated celebrity interactions as inauthentic. K-pop “digital twins” hosting fan conversations came across as hollow – research found that fans felt these AI-generated interactions lacked authenticity and did not constitute genuine communication with the artists.

This tells us that fans typically embrace the artificial when it is transparent but reject it when it is masquerading as authentic human connection. The problem isn’t just AI alone, but rather the erosion of boundaries between human and machine-generated intimacy.

The Reckoning We’re Avoiding

Despite all of these massive structural changes, certain human needs persist. The desire for community and collective meaning has been redirected into algorithmic channels that can’t fully satisfy it. Pattern recognition and interpretive skills, the creative impulse to transform rather than passively consume, the hunger for authenticity and genuine connection – all endure, just hidden.

There’s a trade-off: old fandom had close communities, but not a wide audience.

The moment algorithmic control becomes visible enough to resist, users claim they will push back. But these algorithms’ sophistication lie in their invisibility, shaping behavior just below the threshold of conscious awareness. For many young people, algorithmic platforms are the only culture they’ve known.

We’ve normalized a model where platforms profit from this predatory engagement while disclaiming responsibility for it’s consequences – whether we’re talking about harassment campaigns, misinformation spreading, or authentic community erosion.

We’ve accepted algorithmic curation as inevitable rather than examining whose interests it serves. When we can’t distinguish real grassroots movements from algorithmically manufactured trends, our capacity for collective action is compromised.

We must be honest about how algorithms impact us: they change our actions, split groups, make money from our relationships, and shift control. We require rules that acknowledge that platform design has a significant impact on people, rather than treating platforms as passive. We need to rebuild institutions and practices that foster community outside of algorithmic mediation.

Most importantly, we need to decide whether we’ll allow this to define all cultural production going forward, or whether we’ll fight to preserve our fandom spaces where human connection, creative labor, and collaborative effort can flourish on their own terms.

When did you last create something for your fandom community without thinking about views, likes, or shares? What would it take to reclaim that space?