This article explores the evolving phenomenon of science communication in social media and the norms and values that shape these online spaces. We interviewed science communicators on TikTok, one of the most popular social media platforms globally, about using generative AI to produce science-related social media content.

Sascha Schönig

## Introduction {#introduction .wp-block-heading}

Social media platforms such as Facebook, X and Instagram have not only
changed the way we communicate but also how we learn and acquire
knowledge. Platforms have become spaces for disseminating academic
knowledge, where science communicators convey academic knowledge to the
public as "the new type of influencers" (Zhang & Lu, 2023). Unlike
commercial influencers who tell us what to buy, academic influencers
educate us about various topics. On TikTok, for instance, we see videos
about health, climate change, and many other topics in which science
communicators debunk myths about their professions and share research
findings under the hashtags \#ProfessorsOnTikTok, \#LearnOnTikTok, and
\#SciComm (Zeng et al., 2021). While science communicators can be
faculty members who create media content in their free time (Huber &
Quesada Baena, 2023), others may not have an academic profile, and their
professional profiles are diverse.  

## Platformisation of science communication {#platformisation-of-science-communication .wp-block-heading}

Science communication presents science in ways that are understandable
to the public. It can take various forms, such as documentaries,
podcasts, science journalism, blogs, books, and museums (Burns et al.,
2003). With the advent of social media platforms, science communication
has increasingly taken the form of digital visual and textual posts
(Liang et al., 2014). Platforms shape digital content production,
circulation, and distribution, and academic influencers, science
communicators with high numbers of followers on social media, must
adhere to how platforms operate (Poell et al., 2021).

The way platformisation shapes these activities stems from how platform
companies create and capture value. Social media platforms function as
online spaces that primarily sell advertising. Users can join these
activities for "free" in exchange for their data being collected,
stored, and categorised, and then platforms sell user profiles for
targeted advertising. The more users join a platform, the more user
profiles a platform collects, and the more attractive such a platform
becomes to advertisers. This dynamic, in which an extensive user network
attracts even more users, is called network effects, and it is central
to how platforms create and capture value (Rochet & Tirole, 2003;
Rysman, 2009). 

Social media platforms use various strategies to attract users and
increase their engagement. They create digital infrastructures that
allow users to create, edit, and share content, navigate that content
through internal search, hashtags, and keywords, and express their
opinions through likes and comments. Platforms implement algorithmic
ranking to organise and present content in a way that appeals to users.
Our platform of interest, TikTok, uses a complex filtering mechanism
that is not only based on user engagement statistics (the number of
likes, shares, and comments) but also on whether this content falls into
"Trending Topics." TikTok keeps the algorithms behind Trending Topics a
secret, but what is known is that it combines human and algorithmic
content moderators, current global events, and users' content
preferences. If content belongs to one of the trending topics defined by
TikTok, then the platform promotes such content and directs more user
traffic to it (Zhang, 2021). 

Like other influencers, science communicators must follow social media
platforms' rules to make their content attractive and easily
discoverable. Their creative process begins with selecting a topic (in
their case, a piece of scientific knowledge), then staging scenes,
writing scripts, filming, and editing (Abidin, 2016). When choosing a
topic, academic influencers assess the potential of such content to
generate value, which comes in the form of user numbers and engagement,
such as views, comments, and shares (Duffy et al., 2021). TikTok videos
usually have a solid opening statement or "attention-grabbing hook," a
strong visual appeal, and use specific sound effects. Then, to boost
their metrics, academic influencers must remain consistently active on a
platform, for example, by posting regularly and frequently engaging with
others' content.

## Between academic and online content: Values of science communication {#between-academic-and-online-content-values-of-science-communication .wp-block-heading}

Generally, influencers must balance platform guidelines and user
expectations in creating commercially appealing content while
maintaining authenticity and creativity. "Authenticity" is one of the
critical principles of content production on social media, and it
generally refers to genuineness, originality, and realness (Hund, 2023).
"Being authentic" on social media may mean having a unique perspective,
sharing their personalities, and striving to be 'real' (Arriagada &
Bishop, 2022; Hund, 2023; Craig & Cunningham, 2019; Banet-Weiser, 2012).
"Creativity" is yet another principle of content production. Content
creation is seen as inherently creative, and influencers are generally
expected to produce unique, imaginative, and engaging digital material
that presents new ideas or perspectives. Meeting the creativity standard
helps to maintain audience attention on oversaturated social media
platforms (Duffy et al. 2021). 

While social media and academic content production serve different
purposes, scholars and content creators emphasise creativity, novelty,
and originality in their work. Values are fundamental to academic life,
as they shape the ethical framework for academic conduct (Archer, 2008).
Central to this framework are the liberal values of truth, critical
inquiry, intellectual freedom, and knowledge that benefit society
(Winter & Donohue, 2012). The principle of academic integrity is closely
linked to these values and defines ethical guidelines for academic
practices and behaviour (MacFarlane et al., 2014, p. 340). The
International Center for Academic Integrity (ICAI) outlined six
principles of academic integrity: honesty, trust, fairness, respect,
responsibility, and courage (ICAI, 2021). Fundamentally, academic
integrity revolves around maintaining transparent and trustworthy
scholarly work while cultivating an environment that supports learning
and progress. Violations such as plagiarism and cheating are major
breaches of these norms (MacFarlane et al., 2014). 

A more recent approach highlights the importance of authenticity in
academic work, emphasising the significance of "being true to oneself"
rather than imitating others. It portrays academic culture as authentic,
valuing passionate labour, personal freedom, and genuine
self-expression, and views scholarly work as an empowering and
passionate activity (MacFarlane et al., 2014; Canizzo, 2017). 

As novel technologies emerge and transform practices, they can challenge
existing norms and values. The rise of generative AI has raised concerns
among scholars, who argue it undermines fundamental academic values like
honesty, trust, and fairness by facilitating automated plagiarism
without proper citation (Van Rooij, 2023; Cotton et al., 2023; Susnjak &
McIntosh, 2024). Generative AI brings various challenges and
opportunities for cultural production, such as generating ideas, writing
scripts, and creating images from text prompts (Epstein et al., 2024, p.
1110). Others have observed that generative AI will challenge the
cultural norms and values surrounding social media content production
(Content Creator Scholar Network, 2024).

As AI continues to shape the professional practices of academia and
social media content creation, we want to examine how science
communicators respond to this new technology. 

## Asking the experts {#asking-the-experts .wp-block-heading}

Tinca Lukan conducted ten in-depth interviews with social science
communicators on TikTok, whose follower counts ranged from 2,000 to
110,000. The topics covered in their content spanned sociology,
anthropology, and psychology. All science communicators professionally
belong to academia. Only three were master's students, while the
remaining seven were employed at various levels within universities,
ranging from PhD students to full professors. Among the many questions
about their work as science communicators, the researcher asked their
opinions on using generative AI in their creative processes for
producing science-related social media content. It's important to note
that when asked about generative AI, all participants predominantly
referred to ChatGPT.

## Two camps of science communicators and related norms and values {#two-camps-of-science-communicators-and-related-norms-and-values .wp-block-heading}

Science communicators are divided on generative AI. Some reject it,
while others embrace it. Both groups justify their positions based on
their interpretations of 'academic professionalism' and 'academic
integrity,' as well as on the norms of authenticity and creativity that
drive social media content production.

## Strictly against it: "I do not want to plagiarise someone else's work." {#strictly-against-it-i-do-not-want-to-plagiarise-someone-elses-work. .wp-block-heading}

Science communicators who oppose using generative AI have raised several
ethical concerns, including issues of intellectual integrity, the
perpetuation of biases, lack of transparency, and potential quality loss
of their content. 

Generative AI tools like ChatGPT are trained on diverse datasets but do
not disclose the origins of this data. Therefore,  academic content
creators are unable to verify the originality of the generated content
or ensure appropriate attribution to original authors. This conflicts
with established academic practices of proper citation (Entwistle, 2009;
Noble, 2018; Park, 2010). As one participant noted:

*"I don't use and probably never will for a few reasons. Generative AI
has a lot of ethical problems in terms of taking other people's work.
You have to give credit to people. If you are going to say something
similar to someone else, you should say so, like this is coming from
some stuff I saw before. I would hate to be accidentally involved in
something like that, and there's a real possibility of that /.../ I
don't want to be involved in plagiarising somebody else."*

Further, science communicators highlight that generative AI lacks
intellectual depth because it draws from past data. For instance, the
reason for not using AI tools was the perpetuation of gender bias and
stereotypes: 

*"No, I never use it. Instead, in my content and in my academic work, I
talk a lot about gender bias in AI and how it perpetuates it. So no, I
don't want to be involved in reproducing this."*

At the same time, the unoriginal data on which generative AI draws
undermines creativity and personal expression, both of which are
integral to social media content production. Interestingly, while AI can
potentially generate multiple content pieces for social media, some
content creators reject its use in favour of producing fewer but more
creative and thus valuable content:

"*I find it intellectually lacking. It might be good for initial work
but doesn't give you anything creative. In science communication, it's
all about your own personal perspective that people value, your
creativity, and using AI would be like cheapening it. I would rather
make less content than make more that is inauthentic*."

Another creator shared how they attempted to use a prompt to generate a
video script in their style, but AI failed miserably. They realised with
relief, "*I asked ChatGPT to provide me with a video script in my own
style, but it was irrelevant. I'm still useful*." This comment
underscores the central role of human ingenuity and creativity that
cannot be replaced with AI.

## Building bridges: "I like to use AI to bring me back down to the general audience's expectations." {#building-bridges-i-like-to-use-ai-to-bring-me-back-down-to-the-general-audiences-expectations. .wp-block-heading}

Creators who support generative AI argue that it helps them capture
people's attention, stand out in the enormous landscape of digital
content, and make their material more understandable to the general
audience. 

One creator, for instance, uses generative AI to help craft "hooks" ---
questions or statements placed at the start of a video to capture the
audience's attention, often using clickbait and sensationalistic
language, opening the floor to talk about scientific matters. In her
words:

*"I don't use it to produce an entire video, but sometimes I use it for
ideas for hooks. So, to get people's attention, I am using it. You have
to get their attention, and you have to have a good hook. I know my
videos can sound clickbait-y, but once I get people's attention, I can
talk about science and convey ideas they wouldn't necessarily encounter
if my content did not have a good clickbait hook. There are ways it
could be helpful without turning it into a bad thing".*

Using catchy statements as a tool for audience manipulation may
contradict the professional values of science communication, which
emphasise adherence to principles of accuracy, clarity, and ethical
practice (Cahn, 2010). As she explained, hooks are necessary in the
attention economy of social media, and the value of delivering
science-based content after capturing users' attention prevails. 

A different creator shared another reason for incorporating generative
AI into the creative process. He noted that the specialised language
used in academia can be complex for the general public to understand. To
address this, he uses ChatGPT to make his content more accessible and
engaging for a broader audience. The content creator described his
process of using generative AI as follows:

*"I use it to help me outline topics. I asked ChatGPT to give me ten
points about a topic. As academics, we have knowledge, but sometimes we
lose sight of what people do not know. I have been guilty of that, and
you can easily lose an audience. So, I like to use AI to bring me back
down to the general audience's expectations. For example, I made a
prompt, 'assume that I'm starting with zero knowledge about
misinformation in the news media,' and there were several very
foundational points that I would risk just skipping over."*

This approach emphasises the importance of bridging the gap between
academia and the public and highlights AI's instrumental role in
achieving this goal.

## Conclusion {#conclusion .wp-block-heading}

We examined the norms and values surrounding the use and rejection of
generative AI among science communicators on TikTok and offered several
insights and observations.

As a new phenomenon, science communication on social media integrates
the different values and practices of academic work and commercial
content production. Science communicators demonstrate ongoing skepticism
about using AI (at least the current version of generative AI like
ChatGPT) for developing scripts and forming arguments due to its
contradiction to the values of academic integrity, authenticity, and
creativity. Many content creators from academic backgrounds reject AI
because they fear digital transformation may compromise quality, reduce
control, and erode traditional practices.

However, some academic influencers embrace the new technology. They use
AI instrumentally, particularly for editing their content and adapting
it for the context of social media entertainment. By leveraging digital
platforms such as TikTok, communicators extend the reach of their
research findings to diverse audiences, but this requires tailoring
their content for user expectations and a platform commercial
infrastructure. 

As platforms increasingly shape how we work and live, they also impact
the dissemination of science to broader and younger audiences. Further
research is needed to understand science communication on social media,
including the driving norms and the specific knowledge being shared.
Such studies are essential for developing guidelines and best practices
to ensure that science communication remains accurate, ethical, and
effective in the digital age and across online environments.


Introduction   Social media platforms such as Facebook, X and Instagram have not only changed the way we communicate but also how we learn and acquire knowledge. Platforms have become spaces for disseminating academic knowledge, where science communicators convey academic knowledge to the public as “the new type of influencers” (Zhang &amp;

Messaggi di Rogue Scholar

How TikTok Science Communicators Navigate Norms and Values in the Age of Generative AI