Exploring User Perspectives on ChatGPT: Applications,

Perceptions, and Implications for AI-Integrated Education

Reza Hadi Mogavia,b , Chao Dengc , Justin Juho Kima , Pengyuan Zhoud , Young D. Kwone ,
Ahmed Hosny Saleh Metwallyf,b , Ahmed Tlilif , Simone Bassanellig , Antonio Bucchiaroneg ,
Sujit Gujarh , Lennart E. Nackeb , Pan Huia,i,j,∗
a
The Hong Kong University of Science and Technology, Hong Kong SAR
b
University of Waterloo, Canada
c
Accessible Meta Group, United States
arXiv:2305.13114v3 [cs.CY] 25 Nov 2023

d
University of Science and Technology of China, China
e
University of Cambridge, United Kingdom
f
Smart Learning Institute of Beijing Normal University, China
g
Fondazione Bruno Kessler, Italy
h
International Institute of Information Technology (Hyderabad), India
i
The Hong Kong University of Science and Technology (Guangzhou), China
j
University of Helsinki, Finland

Abstract

To foster the development of pedagogically potent and ethically sound AI-integrated

learning landscapes, it is pivotal to critically explore the perceptions and experiences of the
users immersed in these contexts. In this study, we perform a thorough qualitative content
analysis across four key social media platforms. Our goal is to understand the user
experience (UX) and views of early adopters of ChatGPT across different educational
sectors. The results of our research show that ChatGPT is most commonly used in the
domains of higher education, K-12 education, and practical skills training. In social media
dialogues, the topics most frequently associated with ChatGPT are productivity, efficiency,
and ethics. Early adopters’ attitudes towards ChatGPT are multifaceted. On one hand,

⋆
This is the authors’ preprint version of the paper accepted by the Journal of Computers in Human
Behavior: Artificial Humans (doi: https://doi.org/10.1016/j.chbah.2023.100027)
∗
Corresponding author
Email addresses: rhadimogavi@sigchi.org (Reza Hadi Mogavi), cdeng@accessiblemeta.org (Chao
Deng), jjkimab@connect.ust.hk (Justin Juho Kim), pyzhou@ustc.edu.cn (Pengyuan Zhou),
ydk21@cam.ac.uk (Young D. Kwon), ahmed.hosny@bnu.edu.cn (Ahmed Hosny Saleh Metwally),
ahmed.tlili23@yahoo.com (Ahmed Tlili), simone.bassanelli@unitn.it (Simone Bassanelli),
bucchiarone@fbk.eu (Antonio Bucchiarone), sujit.gujar@iiit.ac.in (Sujit Gujar),
lennart.nacke@acm.org (Lennart E. Nacke), panhui@ust.hk (Pan Hui)

Preprint submitted to Elsevier November 28, 2023

some users view it as a transformative tool capable of amplifying student self-efficacy and
learning motivation. On the other hand, there is a degree of apprehension among
concerned users. They worry about a potential overdependence on the AI system, which
they fear might encourage superficial learning habits and erode students’ social and critical
thinking skills. This dichotomy of opinions underscores the complexity of Human-AI
Interaction in educational contexts. Our investigation adds depth to this ongoing discourse,
providing crowd-sourced insights for educators and learners who are considering
incorporating ChatGPT or similar generative AI tools into their pedagogical strategies.
Keywords: Artificial Intelligence (AI), Generative AI, ChatGPT, Education,
Human-Computer Interaction (HCI),, Early Adopters, Social Media, Qualitative Research

1. Introduction

The twenty-first century is characterized by a rapid proliferation of Artificial Intelligence

(AI) technologies, leading to profound transformations across various aspects of our daily
lives, a theme explored in numerous studies (e.g., see [1, 2, 3, 4, 5, 6]). Among these
transformative changes, the most significant is arguably the revolutionary impact on the
educational sector [5, 7].
This impact is best exemplified by the introduction of AI chatbots in education, which
provide instantaneous responses to students’ inquiries, thus serving as readily accessible
learning resources at any time and location [8, 9, 10]. Concurrently, the use of cutting-edge
machine learning algorithms to analyze student data has empowered educators to pinpoint
areas where individual students require additional support, thereby enhancing the efficacy
of teaching strategies [11]. Furthermore, the advent of AI-based learning systems has paved
the way for adaptive learning. These systems, designed to adjust to each student’s learning
pace and style, provide bespoke educational content. Such personalization has
demonstrated its potential in improving learning outcomes and engagement levels
[12, 13, 14].

2
 1 0 0
9 0
8 0

In te r e s t R e la tiv e to P e a k
7 0
6 0
5 0
4 0
3 0
2 0
1 0
0
3

3
3

23
02

02

02

02

02

02
02

02

20
/2

/2

/2

/2

/2
2

/2

2/
1/

5/
15

29

12

19

26
8
1/

1/

2/
2
1/

1/

1/

2/

2/

2/
Figure 1: The weekly search interest in ChatGPT on Google has consistently increased since its inception,
according to Statista [17].

The Chat Generative Pre-trained Transformer (ChatGPT)1 is an advanced chatbot

technology that utilizes Natural Language Processing (NLP) to generate coherent and
contextually-relevant responses to a variety of human inquiries. In accordance with formal
definitions, a chatbot is a “computer program designed to conduct a conversation with a
user through text or speech” [15]. The ability of ChatGPT to create conversational and
engaging learning environments, customized to individual student needs, underscores its
potential for application in educational contexts. This potential is increasingly recognized,
as evidenced by an exponential growth in global interest in recent months, a trend
observable in worldwide Google search statistics (refer to Figure 1).
This surge in interest highlights the promising prospects of ChatGPT for the development
of advanced AI-powered educational systems. However, it simultaneously underscores the
need for a discerning approach when introducing this technology into educational settings.
The caution advised in its usage stems from potential risks that have been identified in
existing research [16]. As such, the broadening interest in ChatGPT should be tempered by
a careful evaluation of these potential risks.

1
https://chat.openai.com/

3
Why Turn to Social Media for Insights? Since its official rollout, ChatGPT has
rapidly accumulated a significant user base [18]. Notably, social media has emerged as a
primary venue for users to share their experiences and viewpoints regarding this AI tool
[19]. These publicly articulated opinions on social media platforms can yield valuable
insights into the usage of ChatGPT [19]. This data enables researchers to delineate the
primary educational applications of ChatGPT and to assess its potential impacts, both
positive and negative, at an incipient stage.
With this background in mind, we are motivated to take the first steps to fill this empirical
research gap systematically by conducting a social media content analysis (SMCA) of four
prominent social media platforms: Twitter 2 , Reddit 3 , YouTube 4 , and LinkedIn 5 .
SMCA is a popular research approach in the field of Human-Computer Interaction (HCI)
that allows researchers to capture people’s reflective responses [20, 21] to an emerging
technology on a demographically large scale, which can be otherwise quite difficult to
establish through interviews, surveys, or other intrusive research methods [20, 22, 23].
Research Questions. More specifically, our research in this paper is guided by two main
research questions:

• RQ1: What are the predominant educational applications of ChatGPT as reported

by early adopters?

• RQ2: What are the initial perceptions of adopters within the education sector
towards ChatGPT, and how do they evaluate its impact on the educational process?
(Is ChatGPT a blessing or a curse?)

In our research, we consider the perspectives of various early adopter groups, including
researchers, educators, learners, parents, and general users (unspecified roles), all of whom

2
https://twitter.com/
3
https://www.reddit.com/
4
https://www.youtube.com/
5
https://www.linkedin.com/

4
could potentially play a critical role in improving our understanding of the development,
implementation, and use of educational technology [24, 19, 25, 26]. We posit that such a
comprehensive investigation can inform the inclusive and responsible advancement of
ChatGPT in education, especially considering its nascent stage of adoption and
development. In what follows, we outline some of the primary findings of our research:
ChatGPT in Education. Utilizing an inductive coding approach, this study scrutinized
approximately 150,000 words contained in 6,000 text samples to ascertain the prevalent
applications of ChatGPT across various educational contexts. A noteworthy adoption of
ChatGPT was observed in higher education (24.18%), K-12 education (22.09%), and
practical skills learning (15.28%). Additionally, ChatGPT has found significant use in
professional development (9.61%), entrepreneurship (7.18%), lifelong learning (4.64%), and
popular science education (3.25%). A total of 13.77% of the instances pertained to
unclassified learning areas.
ChatGPT’s versatility of use was demonstrated in its wide array of applications including,
but not limited to, assisting in essay and paper composition, guiding learners through
homework assignments and test preparation, aiding in language acquisition, fostering
career progression, honing communication strategies, improving digital literacy, developing
management competencies, facilitating benchmarking and brainstorming sessions,
supporting decision-making processes, providing microlearning experiences, and simplifying
complex concepts in popular science education. Further details are elaborated in Section 4.
ChatGPT: A Double-Edged Sword. Discussions about ChatGPT on social media
platforms primarily focus on three themes: productivity (73.19%), efficiency (63.02%), and
ethics (48.51%).6 Our research suggests that ChatGPT can have both positive and
negative implications for productivity, efficiency, and ethics.
Pioneer users of ChatGPT in educational contexts have expressed varied perceptions.

6
Readers should note that these percentages are not mutually exclusive; some user quotes may fall into
multiple categories.

5
Specifically, 30.62% reported solely positive experiences, 17.15% offered only negative
feedback, 37.34% expressed mixed feelings, and 14.89% maintained a neutral stance.
On one hand, ChatGPT is seen as a helpful tool that enhances students’ self-efficacy and
learning motivation. This is particularly beneficial for students with special needs and/or
everyone else who might struggle with traditional teaching methodologies. The
conversational style of ChatGPT enriches the learning experience. Its capacity to provide
immediate feedback and personalized learning opportunities fosters a deeper connection
between students and the subject matter, leading to enhanced knowledge retention.
Furthermore, the round-the-clock availability of ChatGPT is a boon for students with
demanding schedules, ensuring they do not miss out on learning opportunities. This is
especially important for students from socioeconomically disadvantaged backgrounds who
may have commitments that conflict with traditional learning schedules.
On the other hand, some concerns exist. There are apprehensions that students might rely
excessively on AI systems, leading to superficial learning, reduced social skills, and
impaired communication abilities. It is also feared that it might hinder the development of
key psychological traits and discourage exploration of alternative viewpoints.
In addition, the deployment of ChatGPT in education raises vital concerns about data
privacy and ethics. Further details are elaborated in Section 4.
Contributions. In summation, this study offers several primary contributions to the field:

• We have developed a comprehensive resource that catalogues the most common

applications of ChatGPT within educational environments.

• To the best of our knowledge, our study is the first large-scale qualitative
investigation based on actual user feedback that explores the viewpoints of early
adopters of ChatGPT in the educational context. By illuminating the diverse array
of perspectives held by early adopters in the education sector, we provide a nuanced
understanding of the positive and negative perceptions surrounding ChatGPT.

6
 • The findings from our investigation reveal that ChatGPT is not designed to supplant
the roles of teachers; instead, it is supposed to function as an auxiliary tool that
assists them in elevating the quality of education. The support from educators is
crucial for contextualizing and authenticating the content produced by ChatGPT.

• Finally, the key findings from our study can guide the design of future studies. These
could include surveys, interviews, and controlled experiments that aim to better
understand the diverse and comprehensive views and experiences of different
stakeholders when incorporating ChatGPT in education.

2. Background and Related Work

2.1. Language Models

The field of natural language processing, also known as computational linguistics, has
made significant progress in recent years, especially with the emergence of large neural
network models trained on massive datasets [27, 28]. Unlike traditional rule-based systems,
these large data-driven models achieve greater precision and efficiency in various NLP tasks
[29, 30], such as language translation [31], sentiment analysis [32], named entity recognition
[33], and text summarization [34], among others.
Deep learning has become the dominant approach for developing NLP models [35]. Among
the most popular NLP architectures today are recurrent neural networks (RNNs) [36],
which include long short-term memory networks (LSTMs) and gated recurrence units
(GRUs) [37, 36], as well as state-of-the-art transformer models [38]. In particular, the
transformer model introduced by Vaswani et al. [39] has been highly influential in the field,
serving as the foundation for many large language models such as the GPT family [38].
Transformers use a self-attention mechanism to process input tokens in parallel, as opposed
to sequentially, which is a key difference from recurrent neural networks (RNNs) [39, 38].
This allows transformers to be more efficient in processing large amounts of text data and
capturing long-range dependencies between words. The self-attention mechanism enables

7
the model to better understand relationships between words regardless of their position in
the text, which improves performance on NLP tasks [38].
Language models have greatly benefitted from the use of pre-training techniques [40].
These techniques involve initially training a model on an extensive dataset before
fine-tuning it for a specific task. Pre-trained models such as BERT (Bidirectional Encoder
Representations from Transformers) [41], BART (Bidirectional and Auto-Regressive
Transformers) [42], XLNet [43], and LLaMa from Facebook/Meta [44] have showcased
significant improvements in performance across various natural language processing tasks.
Most large language models today possess the capability to acquire novel tasks with
minimal examples, even if they are not directly associated with their initial training [45].
This phenomenon is called “few-shot learning,” which denotes the ability of these models
to adapt quickly and efficiently to new tasks. This proficiency is derived from the extensive
general knowledge they acquire during the initial training phase, which can be applied to
novel tasks [45]. This transfer learning phenomenon has been witnessed in many contexts,
including the creation of sophisticated chatbots and language translation models [46, 47].

2.2. GPT Family Models

The GPT family of models, Generative Pre-trained Transformers, introduced by OpenAI,

has experienced significant advancements since the initial release of GPT-1 in 2018
[48, 49, 50, 51]. These powerful LLMs have demonstrated proficiency in a wide range of
NLP tasks and have grown steadily in size and complexity [51, 52].
The pioneering model in the GPT series, GPT-1 [48], was built on the transformer
architecture and boasted 117 million parameters, allowing it to process vast amounts of
textual information. This inaugural model established a solid groundwork for subsequent
advancements in the GPT series by laying the foundation for the exploration and
enhancement of transformer-based language models. Despite representing a noteworthy
milestone in natural language processing, GPT-1 exhibited certain shortcomings. For
instance, the model tended to generate redundant text, particularly in response to prompts

8
that were beyond the purview of its training data. Moreover, GPT-1 struggled to reason in
multiple exchanges within a dialogue and was unable to maintain long-term dependencies
within the textual content. Furthermore, while the model exhibited cohesion and fluency in
short text sequences, these qualities diminished in longer passages.
In February 2019, OpenAI debuted GPT-2 [49], a language model that represented a
notable advancement over its predecessor, boasting 1.5 billion parameters. This model
served as a clear demonstration of the potential for LLMs to consistently generate text that
is both cogent and contextually appropriate. However, it also raised concerns about the
possibility of its malicious misuse (e.g., fake news generation) [53]. Despite its numerous
merits, GPT-2 was not without limitations, as it struggled with tasks that required more
complex reasoning and contextual understanding.
GPT-3 [50], launched in June 2020, established a new standard in the field with an
unparalleled 175 billion parameters. This model demonstrated an impressive ability to
generate human-like text, enticing researchers and industries to explore its applications in
various NLP tasks [54, 55]. The GPT-3.5 version further refined the capabilities of the
model, which was later used for the development of ChatGPT.
The most recent addition to the GPT series, GPT-4 [56], was unveiled on March 14, 2023,
and has received widespread attention for its exceptional performance and multimodal
capabilities. Estimated to have trillions of parameters, GPT-4 can efficiently process both
textual and visual inputs, expanding its potential applications and pushing the limits of
AI-generated content even further. Although currently accessible in a restricted capacity,
GPT-4 marks a significant advancement in the field of large language models.
The evolution of the GPT family of models has had a profound impact on deep learning
and NLP research [57]. With each new iteration, these models have demonstrated
increasingly sophisticated language generation capabilities, opening up new research
avenues and applications in various domains [48, 49, 50, 51, 57].
The GPT family is expected to continue to evolve with the release of GPT-5, anticipated

9
in early 2024. This development is likely to bring about advances in language
understanding, multimodal processing, and task-specific fine-tuning. These improvements
are expected to contribute significantly to the ongoing progress in NLP and AI, driving
transformative changes in research and industry contexts. However, as the capabilities of
the GPT family grow, concerns about potential misuse and ethical implications [58, 57] are
also expected to increase, especially in the context of education. Balancing the benefits and
risks associated with these powerful LLMs will be paramount to ensure their responsible
deployment and adoption.

2.3. AI in Education

AI’s expansive influence across various sectors has had profound impacts on organizations,
societies, and individuals [59]. Dwivedi et al. delved into these implications, examining the
opportunities and challenges arising from generative AI technologies such as ChatGPT in
education, business, and society [59]. They underscored AI’s potential to enhance
“productivity” across diverse sectors, while also drawing attention to issues such as privacy
threats and misinformation.
Building on this, Calisto et al. investigated the acceptance and application of AI systems,
with a particular focus on healthcare settings. They noted that health-focused AI tools like
BreastScreening-AI were favorably received by target users, contributing to reduced
diagnosis times and fewer errors [60, 61]. However, they emphasized that trust and
acceptance from end-users are pivotal for the successful implementation of AI techniques
[62].
In the realm of education, AI’s role has been predominantly about creating personalized,
adaptive learning systems to enhance student experiences [63]. The shift in attention has
gone from replacing to empowering teachers, emphasizing their crucial roles when deploying
AI technologies in education [64]. This is further supported by Shoufan’s research on
ChatGPT’s usefulness for programming tasks, which highlights that AI cannot fully
supplant human intelligence, as evidenced by the unavoidable errors in AI responses [65].

10
A cross-cultural study by Luo et al. provided insights from experts in China and the
United States on the implications of ChatGPT for early education [66]. They concluded
that whether ChatGPT’s impact is beneficial or harmful depends on its implementation
and the balance stakeholders strike between its potential benefits, challenges, and risks.
Their research flags several domains where ChatGPT could pose challenges, including
ethical and social concerns, technical limitations, human oversight, and issues of
accessibility and the digital divide.
Adding to the discourse, Murgia et al. explored the use of ChatGPT in education, with a
particular emphasis on adapting online inquiries to meet different literacy levels [67]. Their
findings indicate that while ChatGPT can adjust its responses to match a 4th-grade level,
there remains room for improvement to achieve optimal readability.
Moreover, in self-guided learning environments where ChatGPT served as a digital mentor,
it demonstrated potential benefits for adult learners in defining educational objectives,
identifying resources, and tracking progress [68]. However, scholars raised concerns about
potential over-dependence on ChatGPT and the ambiguity of AI usage policies in different
educational contexts [68].
In a societal analysis, Kristjan and Siibak drew attention to the tendency of tech
companies to release tools like ChatGPT without fully considering their social implications
[69]. They noted mixed reactions towards ChatGPT’s introduction–some called for a ban
in education, while others perceived it as a “technological fix” demanding a re-evaluation of
educators’ roles and teaching philosophies. They further warned against overlooking the
commercial interests of OpenAI, considering its for-profit status [69].
In conclusion, while AI and tools like ChatGPT appear to hold significant potential for
enhancing learning experiences, concerns about accuracy, potential over-reliance, and
unclarified AI usage policies persist. The current literature lacks a holistic exploration of
AI-empowered environments, particularly from the perspectives of students, parents, and
researchers. Therefore, further research is essential to fully comprehend human-AI

11
interaction in educational settings. Our objective is to provide a comprehensive
understanding of the topic by taking into account the perspectives of a wider spectrum of
stakeholders involved.

3. Research Method

The purpose of this study is to explore early adopters’ experiences and thoughts about
integrating ChatGPT into education by analyzing their social media content on various
platforms, i.e., Twitter, Reddit, YouTube, and LinkedIn. Using social media content
analysis, we can extract valuable themes and helpful insights from social media data on a
larger scale [23], ultimately leading to the identification of the primary applications (RQ1)
and positive and negative perceptions of early adopters’ use of ChatGPT in education
(RQ2).
The remainder of this section provides detailed information on our data collection process,
data description, analysis pipeline, and research ethics to provide a clear and
comprehensive understanding of our qualitative study.
Given the advantages and challenges mentioned above of using ChatGPT-like chatbots in
education, this study sought to identify the broader perspectives of early adopters (i.e.,
researchers, educators, learners, parents, and general users (unspecified roles)), about the
educational applications of ChatGPT to unravel its current deficiencies and strengths and
discuss the potentials of future use.

3.1. Data Collection

For this study, we leveraged the application programming interfaces (APIs) of Twitter [70],
Reddit [71], YouTube [72], and LinkedIn [73] to gather data on the use of ChatGPT in
education. Our data collection on LinkedIn was further enhanced by using the advanced
(and legal) data scraping functions available on Helium (v.3) (similar to [74]).
With over 400 million monthly active users [75], Twitter provides an ideal platform for
screening global users’ opinions on ChatGPT and its educational applications. Reddit,

12
with more than 50 million daily active users [76], provides rich data through its various
subreddits, including those discussing the use of ChatGPT in education. YouTube, with
more than 2.5 billion monthly active users [77], is the world’s largest video-sharing
platform and offers valuable long-form audio and visual data, such as seminars, lectures,
and crash courses discussing the use of ChatGPT in education. Finally, LinkedIn, with
more than 300 million monthly active users [78], is the world’s largest professional
networking platform and provides extensive data on the professional use of ChatGPT in
education. We selected these platforms because of their large user bases, rich data,
accommodating terms of use, and potential to provide valuable insights into our research
questions about the use of ChatGPT in education [22, 23].
Inclusion Criteria. To ensure the relevance of the data collected, we used specific
inclusion criteria. The content had to include the keywords “ChatGPT” AND
(“Education” OR “Learn” OR “Teach” OR “Instruct” OR “Study” OR “School” OR
“University”). We chose these keywords based on their prevalence of use in learning
communities, as determined by Ubersuggest7 (similar to [24]). Our analysis was conducted
with the objective of minimizing the likelihood of overlooking crucial information—to the
best of our ability—through the use of an array of keywords. It should be noted that all of
the data used in our research were analysed in text format, but the scope of the data
collection was restricted to content expressed in English. YouTube data were transcribed
into textual format relying on the transcript feature that is accessible on the platform. In
instances where a transcript was not provided by the content producer, we used the
autogenerated transcript functionality offered by YouTube.
Data sampling. Because studying all social media data is impractical, we followed the
advice of existing literature and collected random samples from highly viewed and liked
social media posts, specifically those in the top 25% (upper quartiles), on Twitter, Reddit,
YouTube, and LinkedIn (refer to [79, 74]). To accomplish this, we employed a stratified

7
https://neilpatel.com/ubersuggest/

13
 Table 1: Alternative Keywords for Early Adopters (in Alphabetical Order)
Early Adopter Roles Alternative Keywords
Researcher Academic, Analyst, Expert, Investigator,
Professional, Scholar, Scientist
Educator Coach, Instructor, Mentor, Professor,
Teacher, Trainer, Tutor
Learner Apprentice, Student, Trainee
Parent Father, Guardian, Mother
General User (unspecified role) N/A

random sampling method, without personalization (see [80]), similar to the method used in
a Computer Supported Cooperative Work (CSCW) study conducted by Fogliato et al. [81].
Our decision to use stratified random sampling was motivated by a desire to secure an
impartial and equitable representation of early adopters’ viewpoints, encompassing
informed researchers, educators, learners, parents, and general users. We established this
by searching for the keywords presented in Table 1 within user profiles and manually
annotating the data. The alternative keywords presented in Table 1 were recommended
using a built-in function designed for similar tasks on the popular qualitative research
application called Atlas.ti [82]. Prior to stratified sampling, the majority of data only
reflected the perspectives of researchers, educators, and general users.
Using this sampling technique, we were able to obtain a diverse but manageable sample of
social media posts. It is essential to note that choosing posts with a high number of views
and likes is justifiable because they are more likely to contain content that resonates with a
larger audience. This is crucial to understanding the dominant attitudes and conversations
on these social media platforms [79].
Exclusion Criteria. To ensure the precision and authenticity of our analysis, we limited
our focus to the experiences of early adopters with real and practical knowledge of
ChatGPT in educational settings, which were manually verified by three co-authors
through user-generated content. In compliance with the established literature, we
intentionally excluded user comments, external links, and image-based content from our

14
analysis to avoid an exponential increase in data volume. Our analysis solely incorporated
comments from the Reddit platform because of its discussion-centric style of
communication (see [83]). Examining comments and external links would have greatly
complicated our efforts to authenticate and analyze user experiences, as these sources
frequently lack nuanced details of their parent posts. Although we recognize that some of
the excluded content may have provided valuable insights, we considered this exclusion
necessary to maintain a more rigorous and systematic approach to data collection and
analysis.
Stopping the Data Collection. As our data collection proceeded in parallel with our
data analysis process, we stopped collecting data once we achieved thematic saturation,
which is the point at which no new information could be extracted from our data. The
decision for thematic saturation was unanimously agreed upon by all data coders involved
in the research, comprising three co-authors. To prevent or reduce the chance of premature
termination of data collection, we intentionally analyzed a minimum of 1,000 pieces of
content from each social media platform, although thematic saturation occurred much
earlier in most cases (i.e., on average = 825.69, SD = 145.66). In addition, we used 500
pieces of content to confirm the accuracy of our thematic saturation as a final measurement.

3.2. Data Description

The data analyzed in this work cover the time period between December 1st, 2022, and
April 22nd, 2023, encompassing the five months following the official public release of
ChatGPT. Our study examines 6,000 text samples totaling roughly 150,000 words to
determine how people commonly used ChatGPT in various learning contexts. Compared to
similar prominent HCI studies, the size of our dataset seems reasonable to provide a broad
view of early users’ opinions (e.g., see [79, 84, 24]).
More specifically, our dataset includes 1,500 randomly selected posts from each of the four
social media platforms (see Section 3.1), with equal numbers from five user groups in
education: researchers (300), educators (300), students (300), parents (300), and general

15
 H ig h e r E d u c a tio n
R e s e a rc h e r

E d u c a to r K - 1 2 E d u c a tio n

L if e lo n g L e a r n in g
L e a rn e r
G e n e r a l ( U n c a te g o r iz e d )

P a re n t P r a c tic a l S k ills L e a r n in g

P o p u la r S c ie n c e E d u c a tio n
P r o f e s s io n a l D e v e lo p m e n t
G e n e ra l U s e r
E n tr e p r e n e u r s h ip

Figure 2: Distribution of user discussions concerning the utility of ChatGPT across diverse educational
settings. ChatGPT is primarily discussed in the context of higher education, K-12 education, and practical
skills learning.

users (300). We used 1,000 posts per platform for the thematic analysis, reserving the
remaining 500 (100 from each user group) to confirm the final saturation of the themes.
Figure 2 displays the distribution of user discussions concerning the utility of ChatGPT
across diverse educational settings.
According to our analysis, most posts (55.27%) came from users in North America
(34.23%) and Northern Europe (21.04%). The rest came from: Oceania (4.10%), Asia
(2.88%), Africa (1.17%), and elsewhere (0.98%). It should be noted that for 35.60% of the
data, the user’s location is either not provided or disallowed for collection by the social
media platform.
The use of ChatGPT in education platforms is a more frequently discussed topic in
wealthier, more developed northern countries. This observation, however, could also be
attributed to regional language differences as well as the dominance of other local social
media platforms in other areas. For example, WeChat, Sina Weibo, and Zhihu are more
popular social media platforms in China [85, 86, 87]. Furthermore, it should be noted that
as of April 25, 2023, ChatGPT is not accessible or has stopped providing services in
countries such as China, Cuba, Iran, Italy, Russia, Syria, Uzbekistan, and Venezuela [88].

16
3.3. Data Analysis

In this study, we used a qualitative research approach, specifically the inductive coding
method, to analyze the collected social media data. We decided to conduct a qualitative
study because this type of research is best suited to capture the complex and nuanced
aspects of users’ reactions and responses to new technologies, trends, and phenomena (see
[23, 74]). The inductive method is particularly suitable for our investigation because of the
apparent lack of prior empirical research on user experience related to the use of ChatGPT
in education. Furthermore, the inductive coding approach enables us to develop themes
from the data without preconceived notions or prior assumptions [23, 74].
We used multiple digital tools to facilitate our coding process. Microsoft Excel helped
organize and manage our data. Atlas.ti 8 , a qualitative data analysis software, allowed for
collaborative human-AI sentiment analysis9 , keyword searching, data coding, and report
10
writing. Jamboard , a virtual whiteboard, allowed brainstorming and visualization of
emerging themes. The combination of these tools enabled a thorough and rigorous analysis
of our data. Figure 3 shows the snapshots of these tools used in our research.
Analysis Pipeline. To identify the overarching themes in our data, we employed Braun
and Clarke’s (2006) [89] six-phase framework (similar to [90]). This involved becoming
familiar with the data, creating initial codes, searching for themes, reviewing them,
consolidating titles and definitions, and finally reporting the themes. To obtain a more
comprehensive understanding of people’s opinions and minimize the chances of
unintentional biases as much as possible, data from all four social media platforms were
analyzed concurrently (similar to [24]).
Three authors (henceforth: coders) conducted the thematic analysis through an iterative
coding procedure collaboratively. The third coder was brought in to help resolve any
potential disagreements between the first two coders.

8
https://atlasti.com/
9
The default categories are positive, negative, and neutral
10
https://jamboard.google.com/

17
 (a) Microsoft Excel (b) Coding with Atlas.ti

(c) Jameboard Discussion Board

Figure 3: Snapshots of tools used during the coding process

In the initial phase of coding, each of the three coders independently analyzed the data and
generated an initial set of codes. The coders then met remotely via video conference using
the Zoom application to review the codes identified individually, discuss any disagreements
in coding, and arrive at a consensus on the initial set of codes. From this initial meeting,
we compiled a codebook containing a summary of the codes and examples of data extracts
(quotes) for each code.
We went through the independent coding process, followed by group discussion and
consensus, a total of five times over two months. With each iteration, we refined the codes
by collapsing some codes and separating others into distinct concepts. We also identified
emerging themes throughout the codes. After the fifth round of coding and discussion, we
arrived at a final codebook that contained the most important information in the data and
answered our research questions.
As a final step in the thematic analysis process, we used the Fleiss’ Kappa measure to

18
estimate and report the level of agreement among all coders. We calculated the Kappa
value using the reliability analysis tool in the IBM SPSS Statistics 11 application. Our
study produced a Kappa score of κ = 0.87, indicating almost unanimous agreement among
our team members on the final themes identified [74].

3.4. Research Ethics

In this study, we adhered to established ethical guidelines for collecting and analyzing data
from social media platforms [91, 92, 74]. Before data collection, we obtained ethical
approval from our institution’s research ethics board to conduct this study. Furthermore,
we obtained proper Application Programming Interface (API) keys from the platforms to
legally access their data without overburdening their systems. We meticulously followed
each platform’s terms of service to ensure compliance with platform policies and user
privacy. Furthermore, we only collected publicly available data and anonymized them by
removing any personally identifying information. Furthermore, it should be noted that all
coders in this study have received certified training in human subjects protection
regulations, policies, and ethical issues.

4. Findings

The study’s findings are divided into two sections based on the research questions posed:
(1) uncovering ChatGPT’s most frequent applications in education, and (2) exploring
different perspectives on ChatGPT’s utility in education. The main themes within each
section are presented as bullet points. Quotes from the content analysis are indicated in
italics, and the sources of different quotes and the positions (roles) of each commenter are
reported while ensuring anonymity. Moreover, the percentage values indicate the frequency
with which a topic appears during the content analysis.

11
https://www.ibm.com/products/spss-statistics

19
 (a) Higher Education (b) K-12 Education (c) Practical Skills Learn- (d) Professional Develop-
ing ment

(e) Entrepreneurship (f) Lifelong Learning (g) Popular Science Edu- (h) General Learning Do-
cation main (Uncategorized)

Figure 4: The word clouds in this figure depict the top fifty applications of ChatGPT within various learning
domains, sourced from social media platforms.

4.1. RQ1: Exploring the Multifaceted Uses of ChatGPT in Education

Our research reveals that ChatGPT is most frequently used in three distinct educational
settings: higher education (24.18%), K-12 education (22.09%), and skills training (15.28%).
Other areas where ChatGPT has been used include professional development (9.61%),
entrepreneurship education (7.18%), lifelong learning (4.64%), and popular science
education (3.25%). Furthermore, 13.77% of the instances are related to other learning
domains that are not explicitly categorized. To visually represent the widespread use of
ChatGPT in education, Figure 4 displays the word clouds from social media-generated
content, showcasing the top 50 applications of ChatGPT within each learning domain. In
the following sections, we explore each category in more detail.

4.1.1. Higher Education (24.18%)

In the context of higher education, which refers to tertiary education following secondary
education, typically in a college or university setting, ChatGPT has shown significant
potential in helping researchers, educators, and students.
• Content Creation and Editing. The most prevalent use of ChatGPT in higher education

20
is content creation and editing, accounting for 78.11% of the reported applications.
Educators, students, and researchers have used ChatGPT to write and proofread research
papers, prepare lecture notes, create class presentation slides, and compose literature
reviews. One professor shared her experience on YouTube, stating:
“Simply enter the article or book chapter that you want to present in your next meeting or
class and then presto! You have a nice outline for your slides, and the content can be edited
however and as how many times as you like! ”
• Content Summarization. Furthermore, ChatGPT has proven helpful in summarizing
academic papers, as shared by a Ph.D. student on LinkedIn:
“ChatGPT is awesome for getting a quick rundown of authors’ works. I’m finding it super
helpful lately, and it could be a great way to sift through articles for a lit review and figure
out what’s [relevant] and worth reading and what’s not.”
• Collaborative Data Analysis. The integration of ChatGPT in higher education has not
only streamlined various processes but also inspired new methods of research and data
analysis. A distinguished professor of social science and humanities has proclaimed on their
YouTube channel, “ChatGPT is the beginning of a new era of big data exploration.”
ChatGPT can serve as an AI-based “research assistant for text-based data analysis,
annotation, and sentiment analysis.” A helpful prompt this professor employs to improve
their thematic analysis is asking ChatGPT, “What are other angles this [issue, topic, or
quote] could be analyzed from? ”
• Student Evaluation. Teaching Assistants have found value in using ChatGPT for student
evaluation. One TA shared on Twitter:
“ChatGPT saved me a lot of time and effort while grading 80 C++ homework submissions!
I requested ChatGPT’s assistance in creating a grading [rubric], and it was fantastic. I was
able to review all the submissions, test them more [effectively], and even detect some
instances of plagiarism.”
• Peer-Review Processes. Finally, the peer-review process is another area where ChatGPT

21
 F o r e ig n L a n g u a g e s
S o c ia l S c ie n c e s 2 0 .2 2 % O th e r A r e a s U n s p e c ifie d
U n s p e c ifie d C o o k in g a n d B a k in g P r o g r a m m in g a n d C o d in g O th e r A r e a s
U n s p e c ifie d 1 5 .9 % 3 0 .2 1 % 3 5 .2 %
2 1 .1 1 % 2 .3 6 % 1 7 .6 1 % 1 9 .6 7 %
2 5 .4 7 %
B u s in e s s a n d M a n a g e m e n t
P r o g r a m m in g M a n a g e m e n t
1 1 .7 3 % 1 2 .4 3 % 2 .5 3 %
O th e r S u b je c ts H e a lth
E d u c a tio n
3 .5 4 % 1 4 .2 8 %
4 .0 4 % IT a n d D ig ita l L ite r a c y
S T E M M a th e m a tic s H o m e Im p r o v e m e n t S a le s a n d M a r k e tin g
1 7 .0 8 %
1 1 .6 3 % 1 1 .0 3 % 2 .4 6 % 3 .4 5 %
O th e r M a jo r s
4 .2 2 % U n s p e c ifie d
N a tu r a l S c ie n c e s C u ltu r a l C o m p e te n c e
5 .4 4 % W e b D e s ig n a n d D e v e lo p m e n t
6 .5 9 % 3 .6 1 % D e s ig n T h in k in g
A r ts a n d H u m a n itie s A r t C o m m u n ic a tio n
H e a lth S c ie n c e s 1 4 .0 3 % 7 .2 9 %
7 .8 9 % L a w a n d L e g a l S tu d ie s
1 0 .5 8 % H e a lth 1 0 .8 8 % U s e r In te r fa c e D e s ig n 4 .2 1 %
S o c ia l S tu d ie s C r e a tiv e W r itin g
6 .8 9 % B u s in e s s W r itin g S k ills
8 .5 4 % 7 .3 1 % 5 .7 2 % 7 .8 9 % 6 .9 6 %

(a) Higher Education (b) K-12 Education (c) Practical Skills Learn- (d) Professional Develop-
ing ment
U n s p e c ifie d L a n g u a g e s C o m p u te r A s tr o n o m y
1 5 .0 4 % 2 1 .9 1 % 2 0 .6 2 % 1 9 .5 %
O th e r A r e a s D e s ig n T h in k in g U n s p e c ifie d
2 7 .2 % 1 2 .0 6 % 2 2 .1 7 % U n s p e c ifie d
P e r s o n a l G r o w th T e c h n o lo g y 1 0 .6 5 %
4 .7 5 % 1 6 .2 2 % C h e m is tr y
D ig ita l M a r k e tin g
3 .5 2 %
3 .9 9 % B io lo g y
1 4 .1 8 %
L e a d e r s h ip a n d M a n a g e m e n t F in a n c e P s y c h o lo g y
C r itic a l T h in k in g
6 .0 3 % H is to r y 9 .0 6 % 5 .3 7 %
1 1 .0 5 %
5 .1 % E c o n o m y
H e a lth G e o lo g y 1 1 .4 6 %
B u s in e s s L a w B u s in e s s M o d e lin g O th e r A r e a s
C r e a tiv e D e s ig n a n d In n o v a tio n A r t 8 .2 8 % 7 .2 1 % O th e r A r e a s
6 .4 2 % 9 .6 5 % 6 .0 2 %
8 .5 6 % 6 .4 9 % 7 .4 9 %

(e) Entrepreneurship (f) Lifelong Learning (g) Popular Science Edu-

cation

Figure 5: Distribution of ChatGPT usage across different disciplines, study subjects, and learning areas

has made an impact. A professor of HCI shared his experience on LinkedIn:

“Time is money ... [So] these days, before agreeing to review [a paper], I use ChatGPT to
first get an overview of [the work] and assess its overall quality. A low ChatGPT score
[ChatGPT-based evaluation] will automatically disqualify a paper from my review [list].
ChatGPT does a great job of summing up the key takeaways and contributions of the
research. ... I highly recommend it to busy colleagues and peer reviewers! ”
Early Adopter Majors. According to our findings, the three main subjects in which
ChatGPT has been most frequently utilized by early adopters in higher education are
Social Sciences (15.9%), Business & Management (11.73%), and STEM Education
(Science, Technology, Engineering, and Mathematics) (11.63%). Figure 5 provides a more
comprehensive overview of the most popular fields of study in which ChatGPT is
reportedly used the most.

4.1.2. K-12 Education (22.09%)

In K-12 education, which refers to primary and secondary education, ChatGPT has been
used to enhance teaching and learning experiences for educators, parents, and students.
The following are some of the key applications of ChatGPT in this context:
• Learning Language Skills. ChatGPT has been instrumental in helping students improve

22
their reading, writing, grammar, sentence-building, and conversational skills. By engaging
with ChatGPT, learners can practice their language skills in a low-pressure environment,
enabling them to gain confidence and fluency. An English teacher shared on Twitter: “I’ve
been using ChatGPT to help my students with their vocabulary and grammar. They love the
interactive aspect and it has made a noticeable difference in their writing skills. Highly
recommended! ”
• Personalized Learning Experiences. ChatGPT can be used to create tailored learning
experiences that cater to individual students’ needs and interests. For instance, teachers
can use ChatGPT to construct collaborative story-building exercises that help students
learn creative writing skills in a fun and engaging manner. A middle school teacher shared
their experience on LinkedIn:
“I’ve been incorporating ChatGPT into my creative writing lessons to create personalized
prompts for each student based on their interests and skill level. It has not only helped them
develop a deeper understanding of story structure and character development but also
motivated them to push their creative boundaries. ... The improvement in their writing and
their increased enthusiasm for the subject are truly remarkable.”
Furthermore, ChatGPT can be integrated into educational online games (or game
environments), promoting active learning and knowledge retention. A parent shared their
experience on YouTube:
“ChatGPT has been a great addition to the educational games [on mathplayground] my son
plays online. It acts as a smart and interactive learning assistant, guiding him through
challenging math puzzles and providing real-time feedback on his problem-solving skills. ...
The way ChatGPT adapts to his learning pace and offers customized hints has made
learning more enjoyable and effective for him. He’s now more motivated than ever to
continue [exploring] and mastering new subjects.”
• Supporting Home Learning. ChatGPT has proven to be a valuable asset for supporting
home learning by providing assistance with assignments and helping parents teach complex

23
concepts. By using ChatGPT, parents can access a wealth of information and resources,
enabling them to better support their children’s educational journey. A parent shared in a
tweet: “ChatGPT has made a huge difference in our home learning experience. I feel more
confident helping my daughter with her math homework because it offers step-by-step
guidance and clear explanations. It’s like having a personal tutor at our fingertips! ”
• Supporting Accessibility for All Students. ChatGPT has the potential to empower
students with disabilities and special needs by providing them with accessible learning
resources. For instance, ChatGPT can help students with ADHD
(Attention-Deficit/Hyperactivity Disorder) by offering concise explanations and engaging
content tailored to their learning style. A special education teacher mentioned on LinkedIn:
“ChatGPT has been a game-changer in my classroom, especially for my students with
ADHD. By using ChatGPT, I’ve been able to break down complex topics into shorter, more
digestible segments that hold their attention. Additionally, the interactive nature of the
platform keeps them engaged by providing instant feedback and allowing them to work at
their own pace. The result is a more focused learning experience, leading to better
understanding and retention of the material.”
An educator who works with students with learning disabilities shared their experience in a
Reddit post: “Using ChatGPT, I’ve been able to develop personalized resources that cater
to my students’ unique learning needs. For example, I can create visual aids and simplified
explanations that help them grasp complex concepts more easily. This has made a
significant difference in their comprehension and overall progress. ... By leveraging
ChatGPT’s capabilities, educators can create inclusive learning environments that cater to
the diverse needs of all students.”
Early Adopter Study Subject. Based on the results of our investigation (refer to
Figure 5), early adopters predominantly employed ChatGPT within the K-12 educational
domain, concentrating on three primary subjects: foreign languages (20.22%),
programming (12.43%), and mathematics (11.03%).

24
4.1.3. Practical Skills Learning (15.28%)

In the realm of practical skills learning such as “Do It Yourself (DIY)” communities [93],
ChatGPT has been utilized to support people in acquiring and refining hands-on abilities
in various fields. Key applications of ChatGPT in this context include:
• Problem-solving Assistance. ChatGPT has been influential in assisting learners with
problem-solving tasks, such as debugging code or troubleshooting a malfunctioning device.
By providing step-by-step guidance and engaging in question-and-answer interactions,
ChatGPT enables users to develop a deeper understanding of the issues they face. For
instance, one user mentioned on Reddit: “I was struggling with a coding issue for hours,
and ChatGPT helped me find the bug in no time. The [tool] provided clear instructions and
explanations, making it easy for me to fix the problem and learn from the experience.
• Personal Skill Development. ChatGPT has proven valuable in helping users develop a
wide range of practical skills by providing access to guidance on various projects,
techniques, and tools. Examples include website development, cooking, learning a new
language, and acquiring information about health and personal care. A freelance web
developer shared on Twitter: “ChatGPT has been an amazing resource for me as I’ve been
learning web development. It offers tips, best practices, and even code snippets to help me
polish my skills and create more professional websites.”
• Supporting Children’s Skill Development. Parents and teachers can use ChatGPT to find
age-appropriate projects and resources that help their children develop practical skills in
various areas, such as programming, storytelling, cooking, or crafting. A parent shared on
a Reddit comment: “Thanks to ChatGPT, I’ve found so many awesome projects for my
kids during their school breaks. It gives ideas that match their interests and even explains
how to tackle tricky parts step by step. My kids have built robots, tried coding, and let their
creativity run wild, all while having a blast. I love how ChatGPT makes learning fun and
accessible and helps my family and kids grow their skills and confidence! ”
• Career Exploration. ChatGPT can help users explore different career paths that involve

25
practical skills, providing information about job requirements, necessary qualifications, and
potential opportunities. Users can ask tailored questions like, “What does it take to become
an AI specialist? ” or “How can I excel in a web development career? ” A college student
shared on LinkedIn: “ChatGPT has been an enlightening guide as I navigate the maze of
career paths connected to my major. By shedding light on the practical skills and
qualifications unique to each field, I’ve gained a clearer understanding of my options, ...
empowering me to make well-informed decisions about my time. ChatGPT reveals insights
that often remain hidden beyond the pages of academic textbooks, making it an invaluable
ally in the quest for career exploration and skill development.”
Early Adopter Learning Areas. As shown in Figure 5, our analysis reveals that within
the realm of practical skills learning, early adopters have predominantly harnessed the
power of ChatGPT to enrich their knowledge and expertise in the areas of programming and
coding (17.61%), health and self-care (14.28%), and web design and development (14.03%).

4.1.4. Professional Development (9.61%)

In the context of professional development education, ChatGPT has been employed to

enhance a variety of skills and knowledge across multiple sectors. From improving
communication abilities to facilitating workshop organization, ChatGPT has demonstrated
its value in numerous professional settings.
• Communication Skills. ChatGPT has so far played a noteworthy role in empowering
professionals to refine their written and verbal communication skills, thereby amplifying
their effectiveness in their respective careers. By providing tailored suggestions and
adjustments, ChatGPT guides individuals in the construction of well-organized emails,
reports, proposals, and presentations. A project manager shared their experience on
LinkedIn:
“ChatGPT has changed how I handle communication, giving me confidence in my writing
and presenting skills. With its smart tips and on-the-spot feedback, I can write content
that’s not just clear and to-the-point, but also more engaging and convincing. ... Thanks to

26
[ChatGPT], my communication game has leveled up, and it’s really made a difference in
how I connect with people at work.”
By offering natural language explanations of trends, patterns, and correlations, ChatGPT
simplifies the process of interpreting and communicating data-driven findings. A data
researcher working at Microsoft mentioned on LinkedIn:
“[ChatGPT] not only helps me identify key insights from large datasets but also
communicates those findings in a clear and concise manner that is easily understood by my
colleagues and clients. It’s like having an AI-powered data storytelling partner on my
team! ”
• Digital Literacy. Equipping individuals with digital literacy skills is critical in today’s
technology-driven world. Reportedly, ChatGPT has been employed by some industry
sectors to facilitate learning in various areas of digital literacy, such as software usage, data
analysis and management, cybersecurity, and online communication etiquette for
employees. A digital marketing specialist shared on LinkedIn: “ChatGPT has proven to be
a valuable tool for my professional development. It enables me to rapidly understand new
concepts, while simultaneously enhancing my decision-making abilities for my clients’
benefit. It’s akin to having a personal digital literacy tutor, accessible 24/7 for tailored
guidance and support.”
• Problem-solving. ChatGPT has been used to boost problem-solving skills in various
professional settings. By asking questions and exploring different solutions, users can work
with ChatGPT to understand complex problems and come up with creative ideas. A
business analyst mentioned in a Reddit post: “Lately, when I’m stuck with a tough problem
at work, I turn to ChatGPT for help. It gets me to think outside the box and see other
viewpoints that I might not have thought of. The interactive and collaborative vibe of
ChatGPT has really improved my problem-solving skills and overall job performance and
happiness.”
• Workshop Organization. Additionally, some users have mentioned that ChatGPT has

27
been a helpful tool in organizing and facilitating workshops, training sessions, and
seminars. By creating agendas, activity plans, and resource materials, ChatGPT simplifies
the planning process, letting facilitators concentrate on providing engaging and effective
learning experiences. A corporate trainer shared their experience on YouTube:
“ChatGPT has been quite useful in organizing workshops for my clients. For instance,
while brainstorming activities, it suggested a hands-on group exercise to encourage creative
thinking, which was well-received by participants. It also assisted in creating detailed
agendas and training materials, reducing my preparation time and allowing me to
concentrate on delivering high-quality content. ... The feedback from participants has been
predominantly positive, and I believe a significant portion of that success can be attributed
to the support I’ve received from [ChatGPT].”
Early Adopter Learning Areas. As depicted in Figure 5, our findings demonstrate
that, in the context of professional development, early users have mainly utilized ChatGPT
to enhance their understanding and proficiency in the fields of Information Technology (IT)
and Digital Literacy (17.08%), Design Thinking 12 (7.29%), and Business Writing Skills
(6.96%).

4.1.5. Entrepreneurship Education and Apprenticeship Programs (7.18%)

In the domain of entrepreneurial skills and expertise, ChatGPT has emerged as an

invaluable information resource for entrepreneurs and startups seeking to innovate,
develop, and expand their businesses. Spanning across brainstorming, benchmarking, and
proposal writing, ChatGPT has reportedly played a beneficial role in numerous facets of
the entrepreneurial journey.
• Brainstorming. ChatGPT has demonstrated its usefulness for entrepreneurs by
generating innovative ideas and presenting various viewpoints. It aids the ideation process
by suggesting enhancements, pinpointing potential issues, and fostering creativity. A

12
As a problem-solving framework [94]

28
startup founder shared their experience on YouTube: “With ChatGPT involved, it’s like
having an additional team member who brings a wide range of diverse ideas, enriching [our]
group’s creative thinking. ... ChatGPT has been a helpful addition to our brainstorming
sessions, offering valuable insights and [alternative] approaches that led to positive,
profitable outcomes. For example, [it] once suggested an interesting marketing strategy
targeting a previously neglected customer segment [in the user feedback pool], resulting in
success. ... Although we’ve seen encouraging results, we understand the need for ongoing
evaluation and tracking its performance over time.”
• Benchmarking. ChatGPT can help entrepreneurs and students gain insights into
industry trends, establish benchmarks, and determine essential performance indicators for
comparative analysis. While the information supplied may occasionally be somewhat dated
owing to ChatGPT’s training data constraints, the AI assistant remains capable of
delivering valuable insights into prevailing industry standards and effective benchmarking
practices. A co-founder of a tech startup shared on LinkedIn:
“Using ChatGPT, we’ve been able to identify crucial benchmarks and performance metrics
specific to the e-commerce industry. ... [Among its contributions,] ChatGPT assisted us
discover innovative strategies for reducing cart abandonment rates, which we hadn’t
considered before. While some information may occasionally be less relevant, ChatGPT
consistently provides valuable insights that enhance our benchmarking comparisons,
enabling us to gain a deeper understanding of our market position and make data-driven
strategic decisions with increased confidence.”
• Proposal Writing. ChatGPT has been utilized by entrepreneurs and their teams to
develop convincing grant proposals and pitch decks. By offering support in conducting
literature reviews, proposing enhancements in structure, and recommending pertinent
(though possibly not the most recent) references, ChatGPT assists users in crafting
compelling content that appeals to potential investors and stakeholders. A grant writer for
a social enterprise shared their experience on YouTube: “It’s awesome to find tools like

29
ChatGPT that make grant proposal writing easier. It’s great at finding relevant literature
and organizing content, which saves a ton of effort. But its real strength is creating a
persuasive language that makes proposals shine. This is key for grant writing, as you gotta
show why your project deserves funding. With ChatGPT, you can be sure your proposals
have a better shot at success, ... even when writing in a language other than English.”
The wide-ranging applications of ChatGPT in entrepreneurship education and
apprenticeship programs emphasize its adaptability and capacity to empower future
entrepreneurs and innovators. This, in turn, cultivates a more vibrant and prosperous
entrepreneurial ecosystem.
Early Adopter Learning Areas. As illustrated in Figure 5, our research shows that
within entrepreneurship education and apprenticeship programs, early adopters
predominantly use ChatGPT to improve their knowledge and skills in Design Thinking
(12.06%), Critical Thinking (11.05%), and Business Modeling (9.65%).

4.1.6. Lifelong Learning (4.64%)

In the context of lifelong learning, ChatGPT has emerged as a valuable tool that supports
self-directed learning by offering a diverse range of resources and features customized to
individual needs. These features include facilitating microlearning experiences,
recommending customized learning paths, and providing ongoing mentoring support.
• Microlearning. ChatGPT excels at delivering engaging, bite-sized learning experiences
through interactive question-and-answer sessions, concise lessons, and clear explanations of
a wide array of phenomena. This targeted approach enables learners to easily grasp and
retain information, making it particularly beneficial for those pressed for time or looking to
deepen their understanding of specific topics.
A YouTuber, who identifies himself as a lifelong learner, shared his enthusiasm for
ChatGPT on his channel: “As someone constantly juggling tasks, I’ve found that ChatGPT
is exceptional at quickly explaining various subjects, from Photoshop techniques and video
editing tips to makeup tutorials. The ability to ask questions and receive straightforward,

30
focused answers has made the learning experience with ChatGPT far more enjoyable than
sifting through search engine results or leafing through magazines.” He concludes,
“Although ChatGPT may not always offer the most current information or be entirely
accurate, it remains an invaluable resource for broadening one’s general knowledge as well
as delving deeper into specific areas of interest or questions.”
• Learning Path Suggestion. ChatGPT can recommend customized learning paths based on
an individual’s self-declared goals and interests, such as establishing mastery in a
programming language or natural language. By providing guidance on the “necessary
learning requirements, resources, and steps,” ChatGPT enhances the learning experience
and helps users achieve their desired outcomes. A software developer shared their
experience on a tech-oriented Subreddit:
“I was eager to learn [a new programming language] but had no idea where to begin.
ChatGPT showed me the ropes, pointing out key concepts, resources, and steps to
[mastering] the language in a matter of weeks. By following its [tailored] recommendations,
I was able to create a personalized learning plan that met [my needs] and allowed me to
[advance] as a professional at my own pace.”
• Continuous Mentoring. ChatGPT serves as a valuable resource for lifelong learners by
offering comprehensive support, including access to information, resources, and
constructive feedback. Whether it involves refining a written piece, nurturing creative
ideas, managing projects, or drafting research findings, ChatGPT effectively functions as a
consistent mentor. Notably, older adults with limited literacy skills have found ChatGPT
particularly inspiring, as it enables them to continue learning without feeling shy about
asking basic questions. Moreover, several laborers have credited ChatGPT for motivating
them to pursue vocational education, given the platform’s availability and accessibility as a
constant educational companion. This highlights the substantial impact ChatGPT has on
diverse learner profiles, providing tailored and timely support to individuals from various
backgrounds. However, it is widely noted among users that ChatGPT’s inability to

31
remember previous contexts poses a challenge that must be addressed in order to provide
more effective and personalized mentoring for lifelong learners.
A writer seeking to develop their skills with ChatGPT’s assistance shared their experience
on YouTube: “[ChatGPT] provides constructive criticism and actionable advice for
enhancing my work, contributing to my growth as a writer. However, its inability to
remember previous interactions can be tedious, as I often need to re-establish context.
Despite this limitation, I believe ChatGPT remains a valuable asset that has had a positive
impact on my writing journey.”
In summary, ChatGPT’s applications in lifelong learning demonstrate its potential to
empower individuals throughout their continuous growth and development.
Early Adopter Learning Areas. As depicted in Figure 5, our study reveals that, in the
realm of lifelong (continuous) learning, early adopters chiefly employ ChatGPT to enhance
their proficiency and expertise in Languages (21.91%), Technology Domain (16.22%), and
Finance (9.06%).

4.1.7. Popular Science Education (3.25%)

Popular science education plays a crucial role in making complex scientific concepts
accessible and engaging for general and often non-expert audiences [95]. It incorporates
various media forms, such as YouTube videos and Medium articles, to effectively
communicate scientific ideas, including topics like quantum physics, quantum computing,
or simply the water cycle. By simplifying abstract concepts and promoting scientific
literacy, popular science education fosters a greater appreciation for the scientific method
and its contributions to society [95]. According to our findings, ChatGPT has become a
valuable tool in this domain, assisting content creators and educators in achieving their
educational objectives.
• Simplifying Complex Concepts. ChatGPT excels at breaking down intricate scientific
ideas into digestible and more comprehensible explanations. By providing analogies,
examples, and concise definitions, it enhances the overall understanding of scientific

32
concepts for a broader audience. A science communicator shared their experience on
LinkedIn: “Using ChatGPT, I recently explained gravitational waves to a group of public
audience members by comparing them to ripples in a pond. This comparison struck a chord
with my audience, piqued their curiosity, and made the topic appear more accessible and
engaging. ChatGPT’s potential for making complex and abstract ideas easier to understand
while maintaining a high level of accuracy has been an incredible boon for communicating
complicated scientific concepts in our online seminars and classes.”
• Storytelling for Education. In addition to simplifying complex concepts, effective popular
science education often relies on storytelling to captivate and inform audiences.
ChatGPT’s natural language processing capabilities make it well-suited to crafting
engaging narratives that weave together scientific concepts and real-world applications. A
YouTube science educator shared their thoughts on the platform: ”Incorporating ChatGPT
into my video creation process has not only helped me simplify difficult topics but also craft
compelling stories that hold my audience’s attention. The combination of clear explanations
and engaging narratives has significantly improved the quality of my content and
contributed to an increase in viewership.”
• Scientific Visualization. The visual representation of scientific concepts poses a
significant challenge to scientific content creators. However, it is crucial for enhancing
comprehension and engagement in popular science education. ChatGPT can aid in
generating innovative ideas for graphs, web-based interactive elements, slide designs, and
other visualization techniques. Recently, ChatGPT has been utilized in conjunction with
generative AI tools such as Midjourney 13 and Stable Diffusion 14 , enabling the creation of
smart prompts, and resulting in more customized image generation. These features
effectively help transform complex information into more tangible and accessible visual
formats. A science writer on LinkedIn shared their experience, emphasizing the value of

13
https://www.midjourney.com/home/
14
https://stablediffusionweb.com/

33
ChatGPT in the realm of popular science education: “Working with ChatGPT has allowed
me to come up with engaging visualizations for my [research-based] articles. ... [Just
recently,] in one of our collaborative attempts, it came up with a clever analogy for a
complex research concept in one of my [pieces,] and I turned it into an eye-catching
infographic. ... When I presented it at a public science [exhibition,] people found it both
easy to understand and visually appealing.”
Early Adopter Learning Areas. As illustrated in Figure 5, our research demonstrates
that early adopters in popular science education predominantly utilize ChatGPT to bolster
their knowledge and skills in the fields of Computer Science and Technology (20.62%),
Astronomy (19.50%), and Biology (14.18%).

4.1.8. General Education (Uncategorized) (13.77%)

While many ChatGPT applications can be readily classified into specific educational and
learning domains, certain instances of its use remain uncategorized due to the lack of
(sufficient) context in social media content. Since ChatGPT applications such as creative
and critical content writing and editing, text summarization, and providing helpful learning
suggestions have been previously discussed and covered in earlier sections, this subsection
will focus on three general topics that emerge more prominently within the context of
uncategorized discourses on ChatGPT’s educational applications: customized curriculum
design, student assessment, and fostering collaboration.
• Customized Curriculum Design. ChatGPT has the potential to help educators develop
customized and adaptive curricula. By understanding the specific learning objectives and
desired outcomes in an educational program (or a course of study), ChatGPT can help
connect the dots and recommend customized instructional resources and approaches
tailored to meet different learning goals. However, a human educator would still need to
guide the overall curriculum design process, ensure the curriculum meets key requirements
and standards, and incorporate human judgment in determining the optimal learning paths
and experiences for students. An instructional designer conveyed their experience on

34
YouTube, stating: “AI won’t replace instructional designers, but it’ll definitely change our
role. With systems like ChatGPT, we can now create, or rather draft, curricula that are
personalized, adaptive, and responsive to various learning objectives super quickly. Even so,
human judgment stays crucial. ... [This means] instructional designers will have an even
bigger responsibility as the planners of human-AI collaboration in education.”
• Student Assessment. ChatGPT’s capacity to generate tailored (practice) exams and
homework assignments, covering a broad range of information and skills, provides a useful
resource for educators, parents, and students themselves to monitor learning progress
effectively. An educator remarked on YouTube, “Previously, the idea of designing
personalized tests appeared unattainable; however, that has changed. Utilizing ChatGPT,
I’ve created [assessments] that address multiple [cognitive] levels and incorporate relevant
real-world examples, which helps my students connect with the material, [fostering] a more
authentic learning experience.”
• Facilitating Collaboration. ChatGPT can play a crucial role in fostering effective
communication and collaboration, especially in diverse and multilingual settings. By
providing translation services and facilitating the exchange of ideas, it helps create an
inclusive learning environment where students feel empowered to work together and share
their insights. A user shared on Reddit, “This is what worldwide teamwork for learning
could look like in the future. With ChatGPT’s ability to understand different languages,
obstacles to connecting with each other disappear and opportunities to work together arise.
Students who struggle to say what’s really on their mind can have their ideas translated and
spread, giving them a space to be heard and making co-creating new knowledge a reality.”
Although ChatGPT shows promise across a variety of educational fields, it is important to
note that in the general education category (unspecified), the majority of learning areas,
specifically more than 85%, lack precise definitions or a concentrated focus, resulting in
broader discussions on education. Moving forward, we explore RQ2, examining both the
positive and negative perspectives of using ChatGPT in education to provide a more

35
Table 2: Frequency distribution of people’s perceptions regarding ChatGPT’s impact on productivity, effi-
ciency, and ethical values across various educational settings.
Educational Setting
Higher K-12 Practical Skills Professional Entrepreneurship Lifelong Popular Science General
Education Education Learning Development Education Learning Education (Uncategorized)
10.70% 3.07% 7.44% 3.02% 1.89% 2.11% 2.25% 1.41%
4.12% 8.24% 3.52% 0.56% 1.16% 0.50% 0.50% 2.91%
Productivity
2.91% 7.97% 5.26% 1.59% 3.09% 1.00% 0.36% 5.76%
1.86% 0.47% 2.20% 2.73% 2.98% 1.93% 0.88% 5.46%
5.89% 3.83% 6.32% 3.25% 1.21% 2.38% 1.29% 1.90%
Frequently
Discussed 10.42% 8.62% 2.80% 1.98% 1.61% 0.84% 0.84% 1.00%
Efficiency
Themes 3.83% 5.31% 5.58% 2.11% 5.10% 1.42% 0.68% 4.83%
4.54% 2.30% 2.88% 2.67% 1.71% 0.52% 0.87% 1.32%
1.92% 4.36% 7.59% 3.60% 0.96% 3.16% 1.06% 1.85%
7.55% 10.58% 3.02% 1.95% 1.47% 0.48% 0.48% 0.68%
Ethics
5.87% 6.28% 6.63% 2.12% 6.01% 1.23% 0.27% 5.59%
3.05% 2.37% 3.12% 2.85% 1.61% 0.20% 0.65% 1.30%
Note: ➢ Positive perception ( ), Negative perception ( ), Mixed Perception ( ), Neutral perception ( )
➢ The most dominant perception in each category is highlighted with a distinctive color.

complete picture of its potential benefits and challenges.

4.2. RQ2: Promises and Perils of ChatGPT in Education

Our analysis of social media discussions reveals that productivity (73.19%), efficiency
(63.02%), and ethics (48.51%) are the predominant topics related to the integration of
ChatGPT into education. Individuals have expressed diverse opinions on these aspects
across various social media platforms. Our findings suggest that ChatGPT can potentially
yield both positive and negative outcomes in terms of productivity, efficiency, and ethics.
Early adopters of ChatGPT within the educational sector have reported mixed
perceptions. Specifically, 30.62% expressed exclusively positive views, 17.15% exclusively
negative views, 37.34% mixed views (encompassing both positive and negative aspects),
and 14.89% maintained a neutral stance. Table 2 presents a comprehensive illustration of
the distribution of individuals’ perspectives concerning the implications of ChatGPT on
productivity, efficiency, and adherence to ethical principles within various educational
settings. In the subsequent sections, we explain these diverse perspectives on the
integration of ChatGPT into education in greater detail.

36
4.2.1. The Spectrum of Opinions on Productivity

In the context of this study, productivity refers to the capacity of ChatGPT to enhance the
quantity of education-related and learning outcomes within a specific timeframe. The
spectrum of opinions on productivity can be divided into four primary groups:
• Positive Impacts on Productivity. Our findings reveal that, with the exception of K-12
education, entrepreneurship, and general education (uncategorized), individuals generally
possess a positive perception of ChatGPT’s utility for improving productivity in the
educational domain. According to our findings, these favorable views on productivity
largely stem from ChatGPT’s capability to streamline various educational processes, such
as content creation, editing, and text summarization.
An educator’s post on LinkedIn exemplifies this positive sentiment at its best: “Ever since
using ChatGPT, I’ve created and updated course materials much faster. It’s made me more
productive. For teachers who spend much personal time on admin work, this could free
them to focus on students.”
• Negative Impacts on Productivity. Negative opinions are primarily found in the context of
K-12 education. Teachers and parents express concern that ChatGPT may exacerbate
procrastination, leading students to become increasingly disengaged from their tasks and
less productive. A teacher elucidated this point on YouTube, stating, “Lazy people may
become even lazier and possibly not do anything at all. ... The thought that the task can be
completed at any time can lead to doing nothing at all.”
These concerns can be better understood in light of the well-known theory of
Self-Determination [96], which posits that individuals are more motivated and engaged
when they experience autonomy, competence, and relatedness in their activities. In
educational settings, students who rely heavily on ChatGPT may feel a decreased sense of
autonomy and competence, as the AI tool assumes most of the responsibility for their
learning and academic success. This diminished sense of self-determination could lead to
lower motivation and engagement, making students more passive and less productive in

37
their education.
• Mixed Views on Productivity. In entrepreneurship education and general education
(uncategorized), opinions tend to be mixed. Users acknowledge both the benefits and
drawbacks of using ChatGPT in education, suggesting that its impact on productivity
depends on certain factors, most notably the prompt design and the ability to verify or
organize generated content. Poorly designed prompts and inadequate domain knowledge
may result in time-consuming and disorganized outcomes, which could ultimately hinder
productivity.
One user on YouTube expressed this sentiment the best by sharing, “ChatGPT can be a
great help when it comes to brainstorming ideas or drafting content, potentially speeding up
the process. However, you still need to invest some time refining the output and making
sure it’s accurate. ... While it may not be a magic solution for productivity, it can be a
valuable tool if used properly.”
• Neutral Views on Productivity. Neutral perspectives do not explicitly express a positive
or negative opinion on ChatGPT’s impact on productivity in education. Instead, they
report potential outcomes and call for further research, experimentation, and observation
to determine ChatGPT’s true influence on educational productivity.
An example of a neutral view shared by a researcher on Twitter is: “As AI writing tools
and models like ChatGPT grow more sophisticated, we need more data and research to
understand their full impact on productivity in education.”

4.2.2. The Spectrum of Opinions on Efficiency

In the context of this study, efficiency refers to ChatGPT’s ability to enhance the quality of
educational processes and outcomes, including improvements in learning, expertise, and
skill development. As with productivity, the spectrum of opinions on efficiency can be
divided into four primary groups:
• Positive Impacts on Efficiency. Our findings show that in the educational settings of
practical skills learning, professional development, lifelong learning, and popular science

38
education, most users possess positive perceptions toward the utility of ChatGPT for
improving the efficiency of learning.
These positive views are often attributed to ChatGPT’s capacity to provide customized
feedback, generate relevant examples, and offer quick and simple answers to a wide variety
of questions. These features can help enhance students’ understanding of complex concepts,
facilitate skill development, and support educators in delivering tailored instruction.
The statement of a self-proclaimed lifelong learner on LinkedIn illustrates this positive
sentiment at its best: “I’m already noticing my conversation skills improving. I’ve been
using ChatGPT to learn a new language by discussing interesting everyday topics in my
target language. ... I ask it to act as an expert on a topic I enjoy, and then it uses all the
specific terms, vocabulary, and slang while responding in the language I’m learning. ... It’s
really engaging and keeps me [motivated] because I get to learn all the important words I
need to be fluent in the subject! ”
Another early user on Reddit provides a similar, more detailed experience sharing: “I’ve
been working on a coding project and I ask [ChatGPT] all the dumb questions I’m too
embarrassed to put online. It’s great because it teaches at my own pace and has infinite
patience. I can run solutions by it, ask if they will work, and if not, why not, and if there
are any issues I haven’t foreseen. I use it to brainstorm, I just start typing out ideas and it
gives me input. If I don’t know how to tackle a certain part of the project, I write out the
problem and ask it for suggestions and pros and cons. If I don’t understand a concept, I
can tell it what I think the concept is and ask if I’ve got it right. If it explains something to
me that I don’t fully understand, I can ask it for examples, or I can rephrase it to the AI
and it will see where I went wrong / what I missed. It’s a lot better than people for
answering those kinds of questions because people aren’t always paying attention to what
you’re asking, or sometimes you have trouble communicating your thoughts in a way the
other person can fully understand. ChatGPT has no problems with that.”
• Negative Impacts on Efficiency. Negative opinions regarding ChatGPT are mainly found

39
in the contexts of higher education and K-12 education. Notably, while most early adopters
in higher education believe that ChatGPT can enhance productivity, they argue that it
might not be as effective in terms of learning quality, particularly for advanced,
intellectual, and creative subjects.
In the K-12 context, concerns about efficiency arise due to ChatGPT’s potential to impede
the development of critical thinking and problem-solving skills among students. Some
educators contend that students might depend too heavily on AI-generated responses,
resulting in a lack of deep understanding and an inability to apply knowledge in novel
situations. Excessive use of ChatGPT could also diminish interaction between students and
teachers, which they deem an essential element of the learning process.
A further concern in both higher education and K-12 settings is the possibility of students
not verifying the validity of the information provided by ChatGPT. As AI-generated
content may not always be accurate, students could inadvertently accept incorrect
information as true, negatively impacting their learning quality.
Wrapping up, a student’s comment on Reddit aptly illustrates this negative viewpoint: “I
bought GPT-4 to help me explain and work through problems in my Calculus 1 class, but it
hasn’t been able to solve even basic math equations that involve things like related rates,
linearization, etc. It’s pretty disappointing to be honest. I heard GPT-4 was connected to
Wolfram Alpha now and I thought it would be reliable. It still confidently provides very,
very wrong answers for almost every Calculus 1 problem I ask it to solve.”
• Mixed Views on Efficiency. Consistent with the mixed opinions on productivity, early
adopters also exhibit mixed perspectives regarding the utility of ChatGPT for efficiency in
the context of entrepreneurship education and other general (uncategorized) educational
domains. Users believe that ChatGPT can be both a valuable resource for enhancing
learning experiences and a potential distraction, depending on how it is utilized. An apt
illustration of such opinions can be found in a post by an early adopter on LinkedIn:
“ChatGPT is super helpful for quick answers and tackling tough topics, but it’s not

40
foolproof. That’s why it’s crucial to rely on our own thinking too. For effective learning, we
need to merge the AI’s input with our critical thinking and problem-solving skills. This
approach helps us truly grasp the subject matter and ensures we’re not solely dependent on
ChatGPT.”
Aligning with this perspective, a compelling instance from the applied educational sphere
of entrepreneurship education is shared by an early adopter on YouTube: “As a beginner,
ChatGPT has really helped me learn a lot about the specific language of entrepreneurship
and how to give clearer feedback to potential clients. ... However, it’s important to note that
it doesn’t quite provide a practical understanding of [the concepts it teaches.] So, I think
asking experienced professionals for advice and attending workshops to fully dive into the
business world atmosphere are still necessary and cannot be fully substituted by ChatGPT.”
• Neutral Views on Efficiency. Similar to neutral perspectives on productivity, neutral
views on efficiency do not explicitly express a positive or negative opinion on ChatGPT’s
impact on efficiency in education. Instead, they report potential outcomes and call for
further research, experimentation, and observation to determine ChatGPT’s true influence
on educational efficiency.
An example of a neutral view shared by a researcher on Twitter is: “The outcome of
ChatGPT on student learning hasn’t been conclusively determined. Further research and
input are necessary to understand its actual role and to establish an appropriate balance
between human-AI interactions in learning experiences.”

4.2.3. The Spectrum of Opinions on Ethics

In the context of this study, ethics refers to the moral implications and concerns associated
with the use of ChatGPT in educational processes. Similar to productivity and efficiency,
the spectrum of opinions on ethics can be categorized into four primary groups:
• Positive Impacts on Ethics. In line with our previous findings on productivity and
efficiency, most early ChatGPT adopters in the educational settings of practical skills
learning, professional development, lifelong learning, and popular science education believe

41
that the tool will help refine ethics in the education sector over time.
These proponents base their arguments on several key factors. Our findings indicate that
some early adopters appreciate ChatGPT’s potential to democratize access to education
and foster equitable learning opportunities for students globally. They assert that the AI’s
capacity to deliver personalized learning experiences and provide instantaneous feedback
can help bridge the divide between students from diverse socioeconomic backgrounds and
those with varying learning needs. Among these supporters, some specifically emphasize
that ChatGPT could be particularly beneficial for learners with disabilities or those
requiring special education.
Furthermore, these supporters underscore the role of ChatGPT in cultivating motivation
for self-education among individuals of all ages and at any time, thus empowering learners
to become more independent and take charge of their own educational journey in a more
confident and engaging manner.
The optimistic view is probably best reflected in a tweet from an early adopter employee,
stating: “ChatGPT makes education easy to access, personalized, and open to everyone. It
helps people from different backgrounds and abilities to learn and work together, breaking
down the usual barriers.”
• Negative Impacts on Ethics. Our investigation of ChatGPT’s impact on ethics primarily
highlights concerns in the context of higher education and K-12 education. The top three
areas of ethical concern encompass academic integrity, the spread of misinformation and
fraud, and privacy and data security issues.
One of the most significant ethical issues pertains to ChatGPT’s potential to promote
academic dishonesty. Specifically, ChatGPT’s ability to produce contextually coherent text
enables students to easily plagiarize, paraphrase, or distort content in their writing
assignments and exams. This compromises academic integrity and undermines the
educational value of specific assessment methods. Consequently, instructors face immense
difficulty in accurately assessing students’ genuine understanding and skills.

42
Another ethical concern involves ChatGPT’s inadvertent propagation of misinformation
and fraud. Although AI models like ChatGPT strive to generate relevant content, they
may inadvertently produce deceptive or erroneous information due to training data
limitations or inherent biases. This raises apprehension regarding the dependability of
ChatGPT-generated content and its potential to mislead students, particularly in the
absence of rigorous fact-verification mechanisms.
Finally, incorporating AI tools like ChatGPT into pedagogical practice necessitates a
strong focus on privacy and data security. AI models generally require vast amounts of
data to operate effectively. Some early adopters express concerns that without well-defined
protection policies and limited information about ChatGPT’s inner workings, there is a
risk of exposing personal data and records of tool usage to potential breaches, unintended
applications, or even misuse.
In conclusion, a grad school instructor’s personal experience shared on Reddit best
exemplifies some of the concerns and challenges surrounding the use of AI-generated
content in academia: “Students had a project report milestone to submit and just reading
through the submissions I could see ChatGPT written all over it. My students who are
international students that struggle to communicate with me in emails all started writing
like native speakers in their reports. But that’s not enough proof that they cheated, right?
Then the fun begins. All 15 students are citing papers that don’t exist. Five to six citations
and references and not a single one is a real paper which gave me enough proof to fail them
on the project and accuse them of using ChatGPT. These students account for 1/3 of my
class who are now more or less failing the class half way through the semester. A lot of the
remaining 2/3 of the class are already struggling.”
• Mixed Views on Ethics. In line with the heterogeneous opinions about productivity and
efficiency, the majority of early adopters express mixed views about the impact of
ChatGPT on promoting or destroying ethical values in the educational context of
entrepreneurship and in other (uncategorized) educational settings. A comment from a

43
business educator on YouTube is a good example that illustrates the mixed views on using
ChatGPT: “ChatGPT is an exceptional tool for communication, but it shouldn’t replace
hard work or be used for dishonest acts like intellectual theft. ... This groundbreaking
technology can help professionals who struggle to express their ideas clearly. By using
ChatGPT to generate alternatives or expand on concepts, they can better share their
knowledge with others. ... Sadly, some may misuse ChatGPT for unethical reasons like
plagiarism or taking credit for others’ work. It’s important to remember that the true value
of ChatGPT is to support and improve your own efforts, not to replace them.”
• Neutral Views on Ethics Regarding ethics, neutral views do not explicitly endorse or
criticize ChatGPT’s influence on ethical considerations in education, similar to the
perspectives on productivity and efficiency. Based on our analysis, neutral comments
predominantly take the form of questions, as people want to know what is considered
plagiarism with ChatGPT and what is not, what is ethical and what is not, and what the
boundaries are. In a tweet, for example, a high school student asks, “Is using AI models
like ChatGPT to assist in developing and enhancing essays and research papers considered
an acceptable approach? ”

4.3. Revisiting Ethics in the Context of AI-driven Education

As noted earlier, AI’s potential to democratize education is promising. However, according

to some critics, it raises the concern of a potential homogenization of thought, as AI
models are predominantly trained on widely available data. This could lead to a subtle
standardization of knowledge and thinking patterns under the guise of ’personalization.’
The possibility of academic dishonesty and misinformation dissemination challenges our
traditional notions of ’knowledge’ and ’learning.’ When AI can generate contextually
relevant and coherent text, we need to reconsider what it means to ’know’ something or
’learn’ something. This suggests the need to shift the focus of education from mere
knowledge acquisition to fostering critical thinking, creativity, and emotional
intelligence—areas where AI still falls short.

44
Data privacy and security concerns in AI-integrated education go beyond safeguarding
data. They bring to the fore the power dynamics inherent in data ownership. Questions
arise about who should own the data generated from students’ learning and who should be
responsible for potential misuse. This necessitates a thorough exploration of data rights
and ownership in an AI context.
The perspectives offering mixed and neutral views highlight the complex relationship
between AI and human agency. The ethical conundrum lies in the blurred boundaries
between AI as a tool to extend human capabilities and AI as an autonomous entity capable
of impacting human lives.
Moreover, the biases inherent in AI could reflect our societal prejudices, effectively holding
a mirror up to societal norms [97]. This revelation can serve as a catalyst for critical
discussions about hidden biases in our daily lives and how we can challenge and address
these biases. This realization underscores the importance of maintaining a
“human-in-the-loop” approach in AI systems [98]. By actively involving humans in the
decision-making processes of AI systems, we can ensure that the technology is supervised,
guided, and corrected by human intelligence and ethical judgment.
To sum up this section, the ethical challenges posed by the use of AI in education provide
an opportunity for societal reflection and transformation. Our response to these challenges
will shape not only the future of education but also our society’s trajectory in an AI-driven
world. The ethical considerations of AI use in education merit ongoing, critical, and
nuanced dialogue among all stakeholders.

5. Recommendations

This study conducted an in-depth analysis of social media content from four platforms to
investigate early adopters’ use and perceptions of ChatGPT in education. Our findings
suggest that users utilize ChatGPT for various reasons and perceive it as a valuable
educational tool that can assist students and educators with various time-intensive tasks,

45
such as research, problem-solving, customized feedback and language learning. However,
early adopters also expressed concerns about the potential misuse, including cheating,
overreliance on the AI tool for answers, indolence and superficial learning.
An essential inquiry arising in this context is the optimal utilization of ChatGPT to
enhance and facilitate educational objectives. Based on the collective insights derived from
research questions RQ1 and RQ2, we put forth the subsequent recommendations to
promote the effective incorporation of ChatGPT in educational settings while concurrently
addressing some of the relevant challenges. Figure 6 provides an overview of all
recommendations in a flowchart.

Unsafe to use
Start ChatGPT
No No No No

Do you know the

Have you Do you know
Are you familiar limitations and
Develop your made sure not yoru rights and Safe to use ChatGPT
sub-tasks Yes with the AI Yes Yes drawbacks of Yes
digital literacy to infringe any how well they (Go to next phase)
technology? using the
laws or rules? are protected?
technology?

dysfunctional class

No No
No

Have you considered

Are your learning Do you assume
Identify the learning a space for students ChatGPT friendly
sub-tasks objectives Yes that students will Yes Yes
objectives to interact/collaborate (Go to next phase)
student-centered? use ChatGPT?
with ChatGPT?

dysfunctional class

No No No No

Do you discuss with

Have you specified the Does each rule have Have you
Developing and your students and Safe to use ChatGPT
sub-tasks usage rules as detailed Yes an understandable Yes communicated the Yes Yes
communicating rules colleagues on how to (Go to next phase)
as possible? explanation? rules with students?
improve the rules?

dysfunctional class
No No

Do you evaluate Does your evaluation

Embracing new actual learning by consider the continuous Ready to incorporate
sub-tasks Yes Yes
evaluation methods looking beyond the progress of students into ChatGPT (Go to next phase)
learning artifacts? account?

Unsafe to use
ChatGPT
No No
No

Do you have the expertise/ Do you accept the

Does your prompt and
Designing effective knowledge to validate the responsbility of the content Safe to use ChatGPT
sub-taks content contribute to Yes Yes Yes
prompts and content design of your prompt and you are generating? Are (Go to next phase)
the learning
generated content? you going to cite
objectives?
ChatGPT?

Incorporating reflective
practices to improve
the learning procedures

Figure 6: An overview of recommendations for integrating ChatGPT into an educational setting

46
5.1. Building Digital Literacy for Responsible AI Integration in Education

Digital literacy encompasses the competencies and skills necessary for individuals to
effectively and responsibly engage with complex information ecosystems [99, 100]. In
today’s rapidly evolving educational landscape, digital literacy is essential for educators
and learners to adeptly utilize technology and adopt the appropriate growth mindset [101].
This is particularly significant when interacting with artificial intelligence tools, such as
ChatGPT, where digital literacy is pivotal in identifying and addressing the emerging
ethical, legal and societal implications [99, 102].
To cultivate digital literacy, grasping technology’s multifaceted capabilities and
fundamental mechanics is imperative. Users should have a holistic view of the underlying
principles and operations of data-driven computational models and algorithms to maintain
realistic expectations of tools such as ChatGPT. With this knowledge, users can discern
these technologies’ potential applications and limitations in educational contexts more
effectively. For instance, it is noteworthy that although ChatGPT can produce coherent
and contextually relevant text, it may encounter difficulties with nuanced or highly
specialized subjects and technical inferences. As a result, using ChatGPT as the main
reference for drafting legal documents is generally inadvisable; however, employing
ChatGPT to provide an initial outline or suggestions for legal content under the
supervision of legal professionals might be acceptable. Similarly, seeking medical advice
from ChatGPT is generally inadvisable, but employing ChatGPT to assist patients in
articulating their thoughts under the guidance of medical professionals might be
acceptable. Thus, the bottom line is that users must exercise prudence when assessing the
accuracy and pertinence of the generated content.
AI systems may unintentionally perpetuate and reinforce existing biases in the source
material by analyzing vast but limited datasets [103]. Hence, educators and learners must
develop skills to investigate, identify and address these biases where possible. Applying
such knowledge when integrating ChatGPT into learning platforms (for example, question

47
pool websites, massive open online courses (MOOCs) and virtual learning environments) is
crucial to foster a fair and inclusive learning environment. Some instructors have suggested
that critical thinking, media literacy and the comprehension of the ethical implications of
AI can serve as effective ways to develop these skills.
Digital literacy also encompasses the ethical dimensions of AI usage, including data privacy
and security, academic integrity and the potential for biased or inappropriate content.
Therefore, educators must adopt a proactive approach to ensure students know the
potential risks and ethical implications of using AI tools such as ChatGPT. This involves
discussing appropriate citation practices to avoid plagiarism, understanding the potential
ramifications of sharing personal or sensitive information with AI systems and informing
users of their rights and responsibilities regarding data ownership and usage. By doing so,
learners are empowered to make informed decisions about utilizing and protecting their
information.

5.2. Reimagining Learning Goals and Objectives in the AI Era

In the era of AI, educators must reconsider learning goals and objectives, given the
likelihood that students will use AI-assisted tools such as ChatGPT for learning. New
objectives should focus on developing students’ abilities to collaborate with AI systems,
compare AI responses, and think critically about the information provided. Activities
encouraging students to debate with ChatGPT or analyze its in-depth responses could be
valuable. Additionally, educators must assess the relevance and meaningfulness of their
courses and lessons in an AI-infused learning environment.
As students increasingly rely on AI tools for support, educators must ensure that learning
outcomes harmonize with this evolving landscape. Theoretical frameworks such as active
learning [104], constructivist learning theories [105], and Bloom’s Taxonomy [106] offer
valuable guidance for adapting learning goals and objectives to account for the presence of
AI.
Bloom’s Taxonomy is a renowned learning framework for organizing educational objectives

48
into hierarchical levels of complexity and specificity. The taxonomy comprises the following
six levels: knowledge, comprehension, application, analysis, synthesis and evaluation [106].
By aligning learning goals and objectives with Bloom’s Taxonomy, educators can establish
a structured approach to promoting cognitive development and critical thinking skills in
AI-enhanced learning. Integrating AI tools such as ChatGPT into the learning process can
support objectives across the various levels of Bloom’s Taxonomy:

• Acquiring Factual Knowledge: ChatGPT can help students establish foundational

knowledge in various subjects. However, educators must instruct students on effective
research methods and encourage using multiple sources for accurate understanding.

• Understanding Complex Ideas: ChatGPT can provide explanations, examples and

analogies to support the development of comprehension skills. Herein, educators must
assess and supplement AI-generated explanations for a holistic understanding
(without oversimplification).

• Applying Knowledge to Solve Problems: ChatGPT can assist students in

problem-solving, brainstorming and generating ideas in diverse subjects, fostering
enhanced problem-solving skills and concept application.

• Critically Analyzing AI-Generated Content: Educators must encourage students to

dissect AI-generated output, focusing on structure, assumptions and biases, to
develop critical thinking skills.

• Synthesizing Diverse Information: ChatGPT can help students combine information

from different sources, promoting well-informed, balanced perspectives and a deeper
understanding of knowledge interconnectedness.

• Evaluating AI-Generated Content: By engaging in debates with AI tools or analysing

AI-generated content in depth, students can develop their evaluative reasoning skills,

49
 learning to assess the quality and credibility of information and make informed
judgments.

5.3. Developing and Communicating Rules for AI in Education

Educators must develop and communicate specific rules outlining the appropriate use of
ChatGPT in various settings. By providing well-reasoned explanations for each rule,
teachers can foster a sense of cooperation and understanding among students [107]. For
instance, a rule might state: “Students are encouraged to use ChatGPT as a supplementary
resource to support their learning and research activities, complementing their engagement
with primary sources, diverse perspectives and critical thinking skills, but not as a
substitute for these essential components of learning. This approach ensures that students
do not excessively rely on AI and, in turn, get the space to develop robust cognitive skills
necessary for independent learning.”
By establishing well-defined parameters, educators can ensure that students harness the
potential of AI technologies while minimizing the risks of misuse or overreliance.
Well-established psychological theories, such as the Theory of Planned Behaviour and the
Self-Determination Theory, can support the establishment of these rules and guidelines.
The Theory of Planned Behaviour (TPB) posits that an individual’s intention to perform a
specific behavior is determined by their attitude towards the behavior, subjective norms
and perceived behavioral control [108]. Educators can influence students’ attitudes and
subjective norms regarding the responsible use of AI tools such as ChatGPT by defining
clear rules and guidelines (especially considering the nascent stage of this technology). This
approach can, in turn, shape students’ intentions to use technology responsibly and
effectively. Furthermore, by explaining the rationale behind each rule, educators can foster
a sense of shared understanding and cooperation, strengthening the subjective norms that
govern responsible AI tool usage.
Self-Determination Theory (SDT) is another psychological framework that emphasizes the
importance of intrinsic motivation and autonomy in human behavior [96]. According to

50
SDT, individuals are likelier to engage in a behavior when they perceive it to be
autonomously chosen and intrinsically motivating. Establishing clear rules and guidelines
for AI tool usage and providing well-reasoned explanations can promote a sense of
autonomy and intrinsic motivation among students. Students may be more inclined to
follow the guidelines and adopt responsible behaviors once they understand the purpose
and benefits of responsible AI usage.
In addition to providing clear rules, maintaining open communication with students about
the responsible use of AI tools is crucial. Regular discussions on the potential benefits and
pitfalls of using ChatGPT in learning can promote a deeper understanding of technology’s
role in education. This dialogue can also facilitate the exchange of ideas, feedback and
experiences, enabling educators and students to collaboratively refine best practices for
effectively integrating AI tools into the learning process.
Moreover, educators should consider regularly reviewing and updating the rules and
guidelines to safeguard their continued relevance and effectiveness. As AI technologies
evolve and new insights emerge from research and practice, adapting the guidelines to
reflect these changes will be essential for maintaining an up-to-date and responsible
approach to AI integration in education.

5.4. Embracing New Evaluation Approaches with AI Integration

Focusing on actual learning outcomes rather than solely on the artifacts produced by
students is vital for outlining effective evaluation approaches [20]. In line with the extant
literature, our findings show that most educators concur that continuous assessment
throughout the semester facilitates a more precise evaluation of learning compared to a few
high-stakes exams [109]. With the advent of ChatGPT, educators are increasingly inclined
to assess students primarily within classroom settings.
Project-based assessments, which align with higher levels of thinking in Bloom’s
Taxonomy, are an efficient means of evaluating learning in environments employing AI
tools. These assessments emphasize the application of knowledge and skills in real-world

51
contexts, focusing on the development of higher-order cognitive processes such as analysis,
evaluation and creation. By devising and executing projects that demand critical thinking,
problem-solving and innovative thinking, educators can motivate students to engage with
AI tools responsibly and thoughtfully, fostering the growth of crucial twenty-first-century
competencies.
Moreover, incorporating peer and self-assessment techniques can enrich the evaluation
process by promoting collaboration, reflection and metacognitive skills among students
[110]. These methods empower students to assess their and their peers’ works, providing
opportunities for constructive feedback and fostering a deeper comprehension of the
assessment criteria. Nonetheless, it is crucial to acknowledge that most instructors deem
focused guidance from a supervisor indispensable in such activities. Such guidance enables
students to reflect more swiftly on their tasks, assignments and exams by participating in
peer and self-assessment activities and utilizing AI tools such as ChatGPT. Many educators
argue that this approach results in more meaningful and enduring learning outcomes.
In addition to re-envisioning assessment and evaluation approaches, educators must
address the ethical implications of AI integration within learning environments. This
entails confronting data privacy, algorithmic fairness and potential biases in AI-generated
content. Integrating discussions of these ethical considerations into the curriculum and
assessment practices enables educators to raise awareness and instill a sense of social
responsibility among students when using AI tools for assignments and exams.

5.5. Designing Effective Prompts and Incorporating Reflective Practices in AI-Assisted

Education

The successful integration of AI tools such as ChatGPT in educational settings increasingly

relies on crafting well-structured prompts and incorporating reflective practices.
Thoughtful prompt design enables educators to leverage ChatGPT’s potential to support
instructional goals and enhance student learning. Furthermore, integrating reflective
practices fosters critical thinking about learning, using AI tools, and developing essential

52
skills and competencies.
Designing quality prompts begins with identifying and aligning clear learning objectives
with appropriate AI-supported activities. This process entails evaluating ChatGPT’s
strengths and limitations and considering students’ needs and contexts. Aligning prompts
with learning goals optimizes AI tools’ potential to support and enrich the learning
experience.
Ensuring the accuracy, relevance and suitability of ChatGPT-generated content requires
testing prompts and verifying content. Educators must thoroughly review AI-generated
content and cross-check it with credible sources to maintain quality and reliability.
Encouraging students to cite ChatGPT when using its content promotes accountability and
ethical practices in AI tool usage.
Educators can facilitate reflection by prompting students to consider the following
questions:

• How did ChatGPT contribute to my understanding of the topic?

• What strategies did I employ to verify the accuracy and reliability of the
AI-generated content?

• How can I use AI tools responsibly and ethically to support my learning?

• What challenges did I encounter while using ChatGPT, and how can I overcome
them in the future?

Lastly, educators should present ChatGPT-generated content engagingly, incorporating

sound pedagogy and study approaches. Techniques such as challenging questions,
problem-solving, providing explanations and storytelling foster active engagement, critical
thinking and knowledge retention. For instance, educators can use ChatGPT-generated
content to create case studies, simulations or role-playing activities, allowing students to
apply their knowledge and skills in real-world contexts.

53
6. Limitations

Inherent to any research, our work is subject to certain limitations that must be
acknowledged when evaluating its explanatory power and generalizability. While our study
provides valuable insights into the experiences of early ChatGPT adopters, it encounters
typical challenges associated with qualitative research, such as unintentional researcher
biases (see [74]) and the complexities of generalizing findings beyond the specific context of
the study (see [74, 111, 112]).
A primary limitation concerns the narrow scope and scale of data collection, which focuses
on a particular subset of posts, time frames, and social media platforms. Consequently, our
investigation may not completely capture the myriad perspectives and experiences of
ChatGPT users, necessitating a prudent interpretation of our findings when drawing
broader conclusions about the larger population of early adopters.
Another limitation to acknowledge is that our study does not purport to encompass the
perspectives of every demographic within the early ChatGPT adopters in education.
Instead, it represents the opinions of a specific group of English-speaking users who have
chosen to share their experiences on the social media platforms investigated in this
research. This nuance is vital, as not all users may be inclined to discuss their experiences
and viewpoints online (see [113, 74]).
To address some of these limitations and advance our understanding of user experiences
and viewpoints, future research could employ a more comprehensive methodological
approach. By incorporating techniques such as Delphi studies, interviews, and surveys,
researchers can gather data from a larger and more diverse sample, thus providing a more
robust representation of the perspectives and experiences of early ChatGPT adopters in
education.

7. Future Work

54
As the nascent field of ChatGPT studies progresses, there is a wealth of research
opportunities that warrant further exploration to either corroborate or challenge the
findings of our preliminary investigations. Based on the data collected and scrutinized in
our research on the use of ChatGPT in education on various social media platforms, we
highlight several avenues for future scholarly inquiries.
Regarding ChatGPT’s practical performance, it is also crucial to assess its effectiveness in
relation to specific learning outcomes, such as knowledge retention, skill development, and
other pertinent objectives. Controlled experiments can be employed to facilitate this
evaluation and determine ChatGPT’s overall efficacy within diverse and specialized
educational contexts. Concurrently, examining interpersonal dynamics is essential to gauge
the impact of ChatGPT on teacher-student and student-student relationships, interaction
patterns, and communication modes. This research area should explore ChatGPT’s
potential to promote peer learning and foster collaborative educational environments.
A critical perspective entails examining the longitudinal effects of ChatGPT on learning
outcomes and users’ perceptions. In-depth, longitudinal investigations can provide a
holistic understanding of ChatGPT’s enduring influence on education. Concurrently, a
comparative analysis should be undertaken to evaluate ChatGPT against other AI
chatbots and traditional pedagogical methods. This comparative framework will enable a
comprehensive examination of the unique advantages and challenges associated with
ChatGPT and alternative educational approaches.
The ethical and responsible integration of ChatGPT in educational settings necessitates an
examination of policy development and teacher training. Policymakers, including academic
institutions, must devise and implement regulations to ensure the judicious use of
ChatGPT. Furthermore, it is crucial to evaluate strategies for equipping educators with the
necessary skills and knowledge to seamlessly incorporate ChatGPT into their teaching
practices.
Addressing the varied needs of learners, research should investigate ChatGPT’s ability to

55
cater to culturally and linguistically diverse students, as well as its application in special
education. This line of inquiry should evaluate potential synergies between ChatGPT and
other assistive technologies to support students with special educational needs, learning
disabilities, or other distinct learning challenges.
Finally, it is essential to explore innovative approaches to enrich the learning experience
with ChatGPT. One such approach involves utilizing affective computing to assess
ChatGPT’s ability to identify and respond to the emotional and social dimensions of
learning, thereby catering to learners’ affective needs. Additionally, integrating ChatGPT
with various instructional strategies, such as game-based learning, gamification techniques,
blended learning, and multimodal content delivery, can serve to enhance student
motivation, engagement, and learning outcomes.
Moreover, the potential of combining ChatGPT with metaverse-based classrooms should be
examined. Metaverse-based classrooms (see [114]), utilizing immersive virtual
environments, offer opportunities for transformative pedagogical experiences that foster
deeper learning and collaboration. Investigating the synergy between ChatGPT and
metaverse-enhanced learning environments can provide valuable insights into the
effectiveness of these innovative approaches in promoting a more engaging and enriching
educational experience for learners.

8. Conclusion

In this study, we investigated the adoption and perception of ChatGPT in education by

analyzing qualitative data collected from various social media platforms. Our analysis
revealed that ChatGPT is utilized in diverse settings for multiple purposes, with the most
widespread uses observed in higher education, K-12 education, and practical skills
development.
Among the numerous applications of ChatGPT in higher education, content creation and
editing emerged as the most prevalent, accounting for 78.11% of reported uses. Our

56
findings unveiled a range of promising opportunities for ChatGPT to support learning,
while simultaneously uncovering critical risks associated with overreliance on the tool.
These risks encompass the potential to limit critical thinking and creativity, impede a deep
understanding of subject matter, and foster laziness and passivity.
Drawing upon insights gleaned from the abundance of crowdsourced knowledge on social
media platforms, our study offered recommendations for a more responsible and effective
integration of ChatGPT in educational contexts. Moving forward, it is clear that ongoing
research is crucial for tracking the progress of language models and AI tools like ChatGPT,
as well as assessing their influence on learning outcomes. In accordance with this
perspective, we concluded our paper by outlining an extensive research agenda for future
exploration.

57
References

[1] D. Wang, J. D. Weisz, M. Muller, P. Ram, W. Geyer, C. Dugan, Y. Tausczik, H. Samulowitz,

A. Gray, Human-ai collaboration in data science: Exploring data scientists’ perceptions of automated
ai, Proc. ACM Hum.-Comput. Interact. 3 (CSCW). doi:10.1145/3359313.
URL https://doi.org/10.1145/3359313

[2] M. E. Aktan, Z. Turhan, İ. Dolu, Attitudes and perspectives towards the preferences for artificial
intelligence in psychotherapy, Computers in Human Behavior 133 (2022) 107273.
doi:10.1016/j.chb.2022.107273.
URL https://doi.org/10.1016/j.chb.2022.107273

[3] C.-Y. Wang, J. J. Lin, Utilizing artificial intelligence to support analyzing self-regulated learning: A
preliminary mixed-methods evaluation from a human-centered perspective, Computers in Human
Behavior 144 (2023) 107721. doi:10.1016/j.chb.2023.107721.
URL https://doi.org/10.1016/j.chb.2023.107721

[4] M. J. Ahn, Y.-C. Chen, Artificial intelligence in government: Potentials, challenges, and the future,
in: The 21st Annual International Conference on Digital Government Research, dg.o ’20, Association
for Computing Machinery, New York, NY, USA, 2020, p. 243–252. doi:10.1145/3396956.3398260.
URL https://doi.org/10.1145/3396956.3398260

[5] X. Zhai, X. Chu, C. S. Chai, M. S. Y. Jong, A. Istenic, M. Spector, J.-B. Liu, J. Yuan, Y. Li, A
review of artificial intelligence (AI) in education from 2010 to 2020, Complexity 2021 (2021) 1–18.
doi:10.1155/2021/8812542.
URL https://doi.org/10.1155/2021/8812542

[6] S. C. Hoi, Responsible ai for trusted ai-powered enterprise platforms, in: Proceedings of the Sixteenth
ACM International Conference on Web Search and Data Mining, WSDM ’23, Association for
Computing Machinery, New York, NY, USA, 2023, p. 1277–1278. doi:10.1145/3539597.3575784.
URL https://doi.org/10.1145/3539597.3575784

[7] A. Seldon, O. Abidoye, The fourth education revolution, Legend Press Ltd, 2018.

[8] F. Clarizia, F. Colace, M. Lombardi, F. Pascale, D. Santaniello, Chatbot: An education support

system for student, in: Cyberspace Safety and Security, Springer International Publishing, 2018, pp.

58
 291–302. doi:10.1007/978-3-030-01689-0_23.
URL https://doi.org/10.1007/978-3-030-01689-0_23

[9] G.-J. Hwang, C.-Y. Chang, A review of opportunities and challenges of chatbots in education,
Interactive Learning Environments (2021) 1–14doi:10.1080/10494820.2021.1952615.
URL https://doi.org/10.1080/10494820.2021.1952615

[10] E. Kasthuri, S. Balaji, A chatbot for changing lifestyle in education, in: 2021 Third International
Conference on Intelligent Communication Technologies and Virtual Mobile Networks (ICICV), IEEE,
2021. doi:10.1109/icicv50876.2021.9388633.
URL https://doi.org/10.1109/icicv50876.2021.9388633

[11] A. Zhang, Human computer interaction system for teacher-student interaction model using machine
learning, International Journal of Human–Computer Interaction (2022)
1–12doi:10.1080/10447318.2022.2115645.
URL https://doi.org/10.1080/10447318.2022.2115645

[12] J. Liu, L. Loh, E. Ng, Y. Chen, K. L. Wood, K. H. Lim, Self-evolving adaptive learning for
personalized education, in: Conference Companion Publication of the 2020 on Computer Supported
Cooperative Work and Social Computing, CSCW ’20 Companion, Association for Computing
Machinery, New York, NY, USA, 2020, p. 317–321. doi:10.1145/3406865.3418326.
URL https://doi.org/10.1145/3406865.3418326

[13] R. H. Mogavi, X. Ma, P. Hui, Characterizing student engagement moods for dropout prediction in
question pool websites, Proceedings of the ACM on Human-Computer Interaction 5 (CSCW1) (2021)
1–22. doi:10.1145/3449086.
URL https://doi.org/10.1145/3449086

[14] M. Xia, M. Sun, H. Wei, Q. Chen, Y. Wang, L. Shi, H. Qu, X. Ma, Peerlens: Peer-inspired
interactive learning path planning in online question pool, in: Proceedings of the 2019 CHI
Conference on Human Factors in Computing Systems, CHI ’19, Association for Computing
Machinery, New York, NY, USA, 2019, p. 1–12. doi:10.1145/3290605.3300864.
URL https://doi.org/10.1145/3290605.3300864

[15] A. Chetlen, R. Artrip, B. Drury, A. Arbaiza, M. Moore, Novel use of chatbot technology to educate
patients before breast biopsy, Journal of the American College of Radiology 16 (9) (2019) 1305–1308.

59
 doi:10.1016/j.jacr.2019.05.050.
URL https://doi.org/10.1016/j.jacr.2019.05.050

[16] S. Sok, K. Heng, Chatgpt for education and research: A review of benefits and risks, Available at
SSRN 4378735.

[17] T. Bianchi, Global weekly interest in chatgpt on google search 2022-2023 (March 2023).
URL https://www.statista.com/statistics/1366930/
chatgpt-google-search-weekly-worldwide/?locale=en

[18] D. Ruby, Chatgpt statistics for 2023 (new data + gpt-4 facts) (March 2023).
URL https://www.demandsage.com/chatgpt-statistics/

[19] A. Tlili, B. Shehata, M. A. Adarkwah, A. Bozkurt, D. T. Hickey, R. Huang, B. Agyemang, What if

the devil is my guardian angel: ChatGPT as a case study of using chatbots in education, Smart
Learning Environments 10 (1). doi:10.1186/s40561-023-00237-x.
URL https://doi.org/10.1186/s40561-023-00237-x

[20] R. Hadi Mogavi, B. Guo, Y. Zhang, E.-U. Haq, P. Hui, X. Ma, When gamification spoils your
learning: A qualitative case study of gamification misuse in a language-learning app, in: Proceedings
of the Ninth ACM Conference on Learning @ Scale, L@S ’22, Association for Computing Machinery,
New York, NY, USA, 2022, p. 175–188. doi:10.1145/3491140.3528274.
URL https://doi.org/10.1145/3491140.3528274

[21] C. P. Smith, J. W. Atkinson, D. C. McClelland, J. Veroff, et al., Motivation and personality:

Handbook of thematic content analysis, Cambridge University Press, 1992.

[22] R. Sinha, R. Shibata, A. Patel, J. A. Sternchos, Social media in minimally invasive gynecologic
surgery: What is #trending on instagram?, Journal of Minimally Invasive Gynecology 28 (10) (2021)
1730–1734. doi:10.1016/j.jmig.2021.02.011.
URL https://doi.org/10.1016/j.jmig.2021.02.011

[23] A. Blandford, D. Furniss, S. Makri, Qualitative HCI Research, Springer International Publishing,
2016. doi:10.1007/978-3-031-02217-3.
URL https://doi.org/10.1007/978-3-031-02217-3

[24] R. Hadi Mogavi, Y. Zhao, E. Ul Haq, P. Hui, X. Ma, Student barriers to active learning in
synchronous online classes: Characterization, reflections, and suggestions, in: Proceedings of the

60
 Eighth ACM Conference on Learning @ Scale, L@S ’21, Association for Computing Machinery, New
York, NY, USA, 2021, p. 101–115. doi:10.1145/3430895.3460126.
URL https://doi.org/10.1145/3430895.3460126

[25] A. Tlili, R. Huang, B. Shehata, D. Liu, J. Zhao, A. H. S. Metwally, H. Wang, M. Denden, A. Bozkurt,
L.-H. Lee, D. Beyoglu, F. Altinay, R. C. Sharma, Z. Altinay, Z. Li, J. Liu, F. Ahmad, Y. Hu,
S. Salha, M. Abed, D. Burgos, Is metaverse in education a blessing or a curse: a combined content
and bibliometric analysis, Smart Learning Environments 9 (1). doi:10.1186/s40561-022-00205-x.
URL https://doi.org/10.1186/s40561-022-00205-x

[26] K.-H. Cheng, C.-C. Tsai, The interaction of child-parent shared reading with an augmented reality
(AR) picture book and parents' conceptions of AR learning, British Journal of Educational
Technology 47 (1) (2014) 203–222. doi:10.1111/bjet.12228.
URL https://doi.org/10.1111/bjet.12228

[27] N. Zhong, Z. Qian, X. Zhang, Deep neural network retrieval, in: Proceedings of the 29th ACM
International Conference on Multimedia, ACM, 2021. doi:10.1145/3474085.3475505.
URL https://doi.org/10.1145/3474085.3475505

[28] P. Liu, W. Yuan, J. Fu, Z. Jiang, H. Hayashi, G. Neubig, Pre-train, prompt, and predict: A
systematic survey of prompting methods in natural language processing, ACM Comput. Surv. 55 (9).
doi:10.1145/3560815.
URL https://doi.org/10.1145/3560815

[29] M. Muller, L. B. Chilton, A. Kantosalo, C. P. Martin, G. Walsh, Genaichi: Generative ai and hci, in:
Extended Abstracts of the 2022 CHI Conference on Human Factors in Computing Systems, CHI EA
’22, Association for Computing Machinery, New York, NY, USA, 2022.
doi:10.1145/3491101.3503719.
URL https://doi.org/10.1145/3491101.3503719

[30] A. Esmaeilzadeh, K. Taghva, Text classification using neural network language model (NNLM) and
BERT: An empirical comparison, in: Lecture Notes in Networks and Systems, Springer International
Publishing, 2021, pp. 175–189. doi:10.1007/978-3-030-82199-9_12.
URL https://doi.org/10.1007/978-3-030-82199-9_12

[31] S. Yang, Y. Wang, X. Chu, A survey of deep learning techniques for neural machine translation

61
 (2020). doi:10.48550/ARXIV.2002.07526.
URL https://arxiv.org/abs/2002.07526

[32] L. Zhang, S. Wang, B. Liu, Deep learning for sentiment analysis: A survey, WIREs Data Mining and
Knowledge Discovery 8 (4). doi:10.1002/widm.1253.
URL https://doi.org/10.1002/widm.1253

[33] J. Li, S. Shang, L. Shao, Metaner: Named entity recognition with meta-learning, in: Proceedings of
The Web Conference 2020, WWW ’20, Association for Computing Machinery, New York, NY, USA,
2020, p. 429–440. doi:10.1145/3366423.3380127.
URL https://doi.org/10.1145/3366423.3380127

[34] A. Elsaid, A. Mohammed, L. F. Ibrahim, M. M. Sakre, A comprehensive review of arabic text

summarization, IEEE Access 10 (2022) 38012–38030. doi:10.1109/access.2022.3163292.
URL https://doi.org/10.1109/access.2022.3163292

[35] I. Lauriola, A. Lavelli, F. Aiolli, An introduction to deep learning in natural language processing:
Models, techniques, and tools, Neurocomputing 470 (2022) 443–456.
doi:10.1016/j.neucom.2021.05.103.
URL https://doi.org/10.1016/j.neucom.2021.05.103

[36] K. Baktha, B. K. Tripathy, Investigation of recurrent neural networks in the field of sentiment
analysis, in: 2017 International Conference on Communication and Signal Processing (ICCSP),
IEEE, 2017. doi:10.1109/iccsp.2017.8286763.
URL https://doi.org/10.1109/iccsp.2017.8286763

[37] L. Yao, Y. Guan, An improved LSTM structure for natural language processing, in: 2018 IEEE
International Conference of Safety Produce Informatization (IICSPI), IEEE, 2018.
doi:10.1109/iicspi.2018.8690387.
URL https://doi.org/10.1109/iicspi.2018.8690387

[38] T. Wolf, L. Debut, V. Sanh, J. Chaumond, C. Delangue, A. Moi, P. Cistac, T. Rault, R. Louf,
M. Funtowicz, J. Davison, S. Shleifer, P. von Platen, C. Ma, Y. Jernite, J. Plu, C. Xu, T. L. Scao,
S. Gugger, M. Drame, Q. Lhoest, A. Rush, Transformers: State-of-the-art natural language
processing, in: Proceedings of the 2020 Conference on Empirical Methods in Natural Language
Processing: System Demonstrations, Association for Computational Linguistics, 2020.

62
 doi:10.18653/v1/2020.emnlp-demos.6.
URL https://doi.org/10.18653/v1/2020.emnlp-demos.6

[39] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, I. Polosukhin,

Attention Is All You Need, arXiv:1706.03762 [cs] (Dec. 2017). doi:10.48550/arXiv.1706.03762.
URL http://arxiv.org/abs/1706.03762

[40] L. Dong, N. Yang, W. Wang, F. Wei, X. Liu, Y. Wang, J. Gao, M. Zhou, H.-W. Hon, Unified
language model pre-training for natural language understanding and generation, in: H. Wallach,
H. Larochelle, A. Beygelzimer, F. d'Alché-Buc, E. Fox, R. Garnett (Eds.), Advances in Neural
Information Processing Systems, Vol. 32, Curran Associates, Inc., 2019.
URL https://proceedings.neurips.cc/paper_files/paper/2019/file/
c20bb2d9a50d5ac1f713f8b34d9aac5a-Paper.pdf

[41] J. Devlin, M.-W. Chang, K. Lee, K. Toutanova, Bert: Pre-training of deep bidirectional transformers
for language understanding, arXiv preprint arXiv:1810.04805.

[42] M. Lewis, Y. Liu, N. Goyal, M. Ghazvininejad, A. Mohamed, O. Levy, V. Stoyanov, L. Zettlemoyer,

Bart: Denoising sequence-to-sequence pre-training for natural language generation, translation, and
comprehension, arXiv preprint arXiv:1910.13461.

[43] Z. Yang, Z. Dai, Y. Yang, J. Carbonell, R. R. Salakhutdinov, Q. V. Le, Xlnet: Generalized

autoregressive pretraining for language understanding, in: H. Wallach, H. Larochelle,
A. Beygelzimer, F. d'Alché-Buc, E. Fox, R. Garnett (Eds.), Advances in Neural Information
Processing Systems, Vol. 32, Curran Associates, Inc., 2019.
URL https://proceedings.neurips.cc/paper_files/paper/2019/file/
dc6a7e655d7e5840e66733e9ee67cc69-Paper.pdf

[44] H. Touvron, T. Lavril, G. Izacard, X. Martinet, M.-A. Lachaux, T. Lacroix, B. Rozière, N. Goyal,
E. Hambro, F. Azhar, et al., Llama: Open and efficient foundation language models, arXiv preprint
arXiv:2302.13971.

[45] I. Beltagy, A. Cohan, R. Logan IV, S. Min, S. Singh, Zero-and few-shot nlp with pretrained language
models, in: Proceedings of the 60th Annual Meeting of the Association for Computational
Linguistics: Tutorial Abstracts, 2022, pp. 32–37.

63
[46] A. Kulkarni, A. Shivananda, A. Kulkarni, Building a chatbot using transfer learning, in: Natural
Language Processing Projects, Apress, 2021, pp. 239–255. doi:10.1007/978-1-4842-7386-9_9.
URL https://doi.org/10.1007/978-1-4842-7386-9_9

[47] T. M. Omran, B. T. Sharef, C. Grosan, Y. Li, Transfer learning and sentiment analysis of bahraini
dialects sequential text data using multilingual deep learning approach, Data & Knowledge
Engineering 143 (2023) 102106. doi:10.1016/j.datak.2022.102106.
URL https://doi.org/10.1016/j.datak.2022.102106

[48] A. Radford, K. Narasimhan, T. Salimans, I. Sutskever, et al., Improving language understanding by

generative pre-training.

[49] A. Radford, J. Wu, R. Child, D. Luan, D. Amodei, I. Sutskever, et al., Language models are
unsupervised multitask learners.

[50] T. Brown, B. Mann, N. Ryder, M. Subbiah, J. D. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam,

G. Sastry, A. Askell, S. Agarwal, A. Herbert-Voss, G. Krueger, T. Henighan, R. Child, A. Ramesh,
D. Ziegler, J. Wu, C. Winter, C. Hesse, M. Chen, E. Sigler, M. Litwin, S. Gray, B. Chess, J. Clark,
C. Berner, S. McCandlish, A. Radford, I. Sutskever, D. Amodei, Language models are few-shot
learners, in: H. Larochelle, M. Ranzato, R. Hadsell, M. Balcan, H. Lin (Eds.), Advances in Neural
Information Processing Systems, Vol. 33, Curran Associates, Inc., 2020, pp. 1877–1901.
URL https://proceedings.neurips.cc/paper_files/paper/2020/file/
1457c0d6bfcb4967418bfb8ac142f64a-Paper.pdf

[51] S. Bubeck, V. Chandrasekaran, R. Eldan, J. Gehrke, E. Horvitz, E. Kamar, P. Lee, Y. T. Lee, Y. Li,
S. Lundberg, et al., Sparks of artificial general intelligence: Early experiments with gpt-4, arXiv
preprint arXiv:2303.12712.

[52] Y. Bang, S. Cahyawijaya, N. Lee, W. Dai, D. Su, B. Wilie, H. Lovenia, Z. Ji, T. Yu, W. Chung,
et al., A multitask, multilingual, multimodal evaluation of chatgpt on reasoning, hallucination, and
interactivity, arXiv preprint arXiv:2302.04023.

[53] K. Hao, Openai has released the largest version yet of its fake-news-spewing ai (August 2019).
URL https://www.technologyreview.com/2019/08/29/133218/
openai-released-its-fake-news-ai-gpt-2/

64
[54] R. Dale, GPT-3: What’s it good for?, Natural Language Engineering 27 (1) (2020) 113–118.
doi:10.1017/s1351324920000601.
URL https://doi.org/10.1017/s1351324920000601

[55] M. Abdullah, A. Madain, Y. Jararweh, ChatGPT: Fundamentals, applications and social impacts, in:
2022 Ninth International Conference on Social Networks Analysis, Management and Security
(SNAMS), IEEE, 2022. doi:10.1109/snams58071.2022.10062688.
URL https://doi.org/10.1109/snams58071.2022.10062688

[56] OpenAI, Gpt-4 technical report (2023). arXiv:2303.08774.

[57] C. Zhang, C. Zhang, S. Zheng, Y. Qiao, C. Li, M. Zhang, S. K. Dam, C. M. Thwal, Y. L. Tun, L. L.
Huy, et al., A complete survey on generative ai (aigc): Is chatgpt from gpt-4 to gpt-5 all you need?,
arXiv preprint arXiv:2303.11717.

[58] P. P. Ray, ChatGPT: A comprehensive review on background, applications, key challenges, bias,
ethics, limitations and future scope, Internet of Things and Cyber-Physical Systems 3 (2023)
121–154. doi:10.1016/j.iotcps.2023.04.003.
URL https://doi.org/10.1016/j.iotcps.2023.04.003

[59] Opinion paper: “so what if ChatGPT wrote it?” multidisciplinary perspectives on opportunities,
challenges and implications of generative conversational AI for research, practice and policy,
International Journal of Information Management 71 (2023) 102642.
doi:10.1016/j.ijinfomgt.2023.102642.
URL https://doi.org/10.1016/j.ijinfomgt.2023.102642

[60] F. M. Calisto, C. Santiago, N. Nunes, J. C. Nascimento, BreastScreening-AI: Evaluating medical

intelligent agents for human-AI interactions, Artificial Intelligence in Medicine 127 (2022) 102285.
doi:10.1016/j.artmed.2022.102285.
URL https://doi.org/10.1016/j.artmed.2022.102285

[61] F. M. Calisto, J. Fernandes, M. Morais, C. Santiago, J. M. Abrantes, N. Nunes, J. C. Nascimento,

Assertiveness-based agent communication for a personalized medicine on medical imaging diagnosis,
in: Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, ACM, 2023.
doi:10.1145/3544548.3580682.
URL https://doi.org/10.1145/3544548.3580682

65
[62] F. M. Calisto, N. Nunes, J. C. Nascimento, Modeling adoption of intelligent agents in medical
imaging, International Journal of Human-Computer Studies 168 (2022) 102922.
doi:10.1016/j.ijhcs.2022.102922.
URL https://doi.org/10.1016/j.ijhcs.2022.102922

[63] W. Holmes, I. Tuomi, State of the art and practice in ai in education, European Journal of
Education 57 (4) (2022) 542–570. doi:10.1111/ejed.12533.
URL https://doi.org/10.1111/ejed.12533

[64] H. Khosravi, S. B. Shum, G. Chen, C. Conati, Y.-S. Tsai, J. Kay, S. Knight, R. Martinez-Maldonado,
S. Sadiq, D. Gašević, Explainable artificial intelligence in education, Computers and Education:
Artificial Intelligence 3 (2022) 100074. doi:10.1016/j.caeai.2022.100074.
URL https://doi.org/10.1016/j.caeai.2022.100074

[65] A. Shoufan, Exploring students’ perceptions of ChatGPT: Thematic analysis and follow-up survey,
IEEE Access 11 (2023) 38805–38818. doi:10.1109/access.2023.3268224.
URL https://doi.org/10.1109/access.2023.3268224

[66] W. Luo, H. He, J. Liu, I. R. Berson, M. J. Berson, Y. Zhou, H. Li, Aladdin’s genie or pandora’s box
for early childhood education? experts chat on the roles, challenges, and developments of ChatGPT,
Early Education and Development (2023) 1–18doi:10.1080/10409289.2023.2214181.
URL https://doi.org/10.1080/10409289.2023.2214181

[67] E. Murgia, M. S. Pera, M. Landoni, T. Huibers, Children on chatgpt readability in an educational

context: Myth or opportunity?, in: Adjunct Proceedings of the 31st ACM Conference on User
Modeling, Adaptation and Personalization, UMAP ’23 Adjunct, Association for Computing
Machinery, New York, NY, USA, 2023, p. 311–316. doi:10.1145/3563359.3596996.
URL https://doi.org/10.1145/3563359.3596996

[68] X. Lin, Exploring the role of ChatGPT as a facilitator for motivating self-directed learning among
adult learners, Adult Learningdoi:10.1177/10451595231184928.
URL https://doi.org/10.1177/10451595231184928

[69] K. Kikerpill, A. Siibak, App-hazard disruption: An empirical investigation of media discourses on

ChatGPT in educational contexts, Computers in the Schools (2023)
1–22doi:10.1080/07380569.2023.2244941.
URL https://doi.org/10.1080/07380569.2023.2244941

66
[70] I. Twitter, Twitter api (January 2023).
URL https://developer.twitter.com/en/docs/twitter-api

[71] I. Reddit, Reddit api (January 2023).

URL https://www.reddit.com/dev/api/

[72] I. Google, Youtube data api (March 2023).

URL https://developers.google.com/youtube/v3

[73] I. LinkedIn, Linkedin api products (January 2023).

URL https://developer.linkedin.com/product-catalog

[74] R. Hadi Mogavi, Y. Zhang, E.-U. Haq, Y. Wu, P. Hui, X. Ma, What do users think of promotional
gamification schemes? a qualitative case study in a question answering website, Proc. ACM
Hum.-Comput. Interact. 6 (CSCW2). doi:10.1145/3555124.
URL https://doi.org/10.1145/3555124

[75] D. Ruby, Twitter statistics for marketers in 2023 (users & trends) (March 2023).
URL https://www.demandsage.com/

[76] D. Ruby, Reddit statistics for 2023 (users & traffic data) (March 2023).
URL https://www.demandsage.com/

[77] D. Ruby, Youtube statistics (2023) — trending facts & figures shared! (January 2023).
URL https://www.demandsage.com/

[78] D. Ruby, Important linkedin statistics for 2023 (data & trends) (March 2023).
URL https://www.demandsage.com/

[79] M. Tahaei, K. Vaniea, N. Saphra, Understanding privacy-related questions on stack overflow, in:
Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems, CHI ’20,
Association for Computing Machinery, New York, NY, USA, 2020, p. 1–14.
doi:10.1145/3313831.3376768.
URL https://doi.org/10.1145/3313831.3376768

[80] K. Papadamou, S. Zannettou, J. Blackburn, E. De Cristofaro, G. Stringhini, M. Sirivianos, ”how

over is it?” understanding the incel community on youtube, Proc. ACM Hum.-Comput. Interact.

67
 5 (CSCW2). doi:10.1145/3479556.
URL https://doi.org/10.1145/3479556

[81] R. Fogliato, A. Chouldechova, Z. Lipton, The impact of algorithmic risk assessments on human
predictions and its analysis via crowdsourcing studies, Proc. ACM Hum.-Comput. Interact.
5 (CSCW2). doi:10.1145/3479572.
URL https://doi.org/10.1145/3479572

[82] A. S. S. D. GmbH, A computer-assisted qualitative data analysis software (February 2023).

URL https://atlasti.com

[83] E.-U. Haq, T. Braud, L.-H. Lee, A. K. Vallapuram, Y. Yu, G. Tyson, P. Hui, Short, colorful, and
irreverent! a comparative analysis of new users on wallstreetbets during the gamestop short-squeeze,
in: Companion Proceedings of the Web Conference 2022, WWW ’22, Association for Computing
Machinery, New York, NY, USA, 2022, p. 52–61. doi:10.1145/3487553.3524202.
URL https://doi.org/10.1145/3487553.3524202

[84] J. A. Pater, L. E. Reining, A. D. Miller, T. Toscos, E. D. Mynatt, ”notjustgirls”: Exploring

male-related eating disordered content across social media platforms, in: Proceedings of the 2019
CHI Conference on Human Factors in Computing Systems, CHI ’19, Association for Computing
Machinery, New York, NY, USA, 2019, p. 1–13. doi:10.1145/3290605.3300881.
URL https://doi.org/10.1145/3290605.3300881

[85] Y. Chen, K.-R. Wang, W. Xu, Y. Huang, Exploring commenting behavior in the covid-19 super-topic
on weibo, in: Extended Abstracts of the 2021 CHI Conference on Human Factors in Computing
Systems, CHI EA ’21, Association for Computing Machinery, New York, NY, USA, 2021.
doi:10.1145/3411763.3451717.
URL https://doi.org/10.1145/3411763.3451717

[86] J. Gong, Z. Yao, Z. Lu, Q. Ding, y. zhang, L. Zhang, Q. Wang, All in one group: Current practices,
lessons and challenges of chinese home-school communication in im group chat, in: Proceedings of
the 2021 CHI Conference on Human Factors in Computing Systems, CHI ’21, Association for
Computing Machinery, New York, NY, USA, 2021. doi:10.1145/3411764.3445436.
URL https://doi.org/10.1145/3411764.3445436

[87] S. Ao, W. Junchi, L. Ke, S. Yuxin, Y. Bingxiang, H. Zhisheng, Analysis of needs and availability of
information on depression based on the zhihu q&a community, in: Proceedings of the 2nd

68
 International Symposium on Artificial Intelligence for Medicine Sciences, ISAIMS ’21, Association for
Computing Machinery, New York, NY, USA, 2021, p. 551–556. doi:10.1145/3500931.3501025.
URL https://doi.org/10.1145/3500931.3501025

[88] S. Conroy, What countries is chat gpt available & not available in? (April 2023).
URL https://www.wepc.com/tips/what-countries-is-chat-gpt-unavailable/

[89] V. Braun, V. Clarke, Using thematic analysis in psychology, Qualitative Research in Psychology
3 (2) (2006) 77–101. doi:10.1191/1478088706qp063oa.
URL https://doi.org/10.1191/1478088706qp063oa

[90] J. Jeon, S. Lee, Large language models in education: A focus on the complementary relationship
between human teachers and ChatGPT, Education and Information
Technologiesdoi:10.1007/s10639-023-11834-1.
URL https://doi.org/10.1007/s10639-023-11834-1

[91] GDPR, The european union’s general data protection regulation (April 2018).
URL https://gdpr-info.eu/

[92] A. of Internet Researchers (AoIR), Internet research: Ethical guidelines 3.0 (April 2019).
URL https://aoir.org/ethics/

[93] B. Lin, E. M. Huang, Reuse: Promoting repurposing through an online diy community, in: CHI ’10
Extended Abstracts on Human Factors in Computing Systems, CHI EA ’10, Association for
Computing Machinery, New York, NY, USA, 2010, p. 4177–4182. doi:10.1145/1753846.1754122.
URL https://doi-org.lib.ezproxy.hkust.edu.hk/10.1145/1753846.1754122

[94] A. L. Culén, A. Følstad, Innovation in hci: What can we learn from design thinking?, in:
Proceedings of the 8th nordic conference on human-computer interaction: Fun, fast, foundational,
2014, pp. 849–852.

[95] H. Xia, H. X. Ng, Z. Chen, J. Hollan, Millions and billions of views: Understanding popular science
and knowledge communication on video-sharing platforms, in: Proceedings of the Ninth ACM
Conference on Learning @ Scale, L@S ’22, Association for Computing Machinery, New York, NY,
USA, 2022, p. 163–174. doi:10.1145/3491140.3528279.
URL https://doi.org/10.1145/3491140.3528279

69
 [96] A. Tyack, E. D. Mekler, Self-determination theory in hci games research: Current uses and open
questions, in: Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems,
CHI ’20, Association for Computing Machinery, New York, NY, USA, 2020, p. 1–22.
doi:10.1145/3313831.3376723.
URL https://doi.org/10.1145/3313831.3376723

[97] B. Foroughi, M. G. Senali, M. Iranmanesh, A. Khanfar, M. Ghobakhloo, N. Annamalai,

B. Naghmeh-Abbaspour, Determinants of intention to use ChatGPT for educational purposes:
Findings from PLS-SEM and fsQCA, International Journal of Human–Computer Interaction (2023)
1–20doi:10.1080/10447318.2023.2226495.
URL https://doi.org/10.1080/10447318.2023.2226495

[98] F. M. Calisto, C. Santiago, N. Nunes, J. C. Nascimento, Introduction of human-centric AI assistant

to aid radiologists for multimodal breast image classification, International Journal of
Human-Computer Studies 150 (2021) 102607. doi:10.1016/j.ijhcs.2021.102607.
URL https://doi.org/10.1016/j.ijhcs.2021.102607

[99] H. Tinmaz, Y.-T. Lee, M. Fanea-Ivanovici, H. Baber, A systematic review on digital literacy, Smart
Learning Environments 9 (1). doi:10.1186/s40561-022-00204-y.
URL https://doi.org/10.1186/s40561-022-00204-y

[100] E. Yi-No Kang, D.-R. Chen, Y.-Y. Chen, Associations between literacy and attitudes toward
artificial intelligence–assisted medical consultations: The mediating role of perceived distrust and
efficiency of artificial intelligence, Computers in Human Behavior 139 (2023) 107529.
doi:https://doi.org/10.1016/j.chb.2022.107529.
URL https://www.sciencedirect.com/science/article/pii/S0747563222003491

[101] Q. Liang, J. de la Torre, N. Law, Do background characteristics matter in children's mastery of

digital literacy? a cognitive diagnosis model analysis, Computers in Human Behavior 122 (2021)
106850. doi:10.1016/j.chb.2021.106850.
URL https://doi.org/10.1016/j.chb.2021.106850

[102] W. Yang, Artificial intelligence education for young children: Why, what, and how in curriculum
design and implementation, Computers and Education: Artificial Intelligence 3 (2022) 100061.
doi:10.1016/j.caeai.2022.100061.
URL https://doi.org/10.1016/j.caeai.2022.100061

70
[103] J. W. Rae, S. Borgeaud, T. Cai, K. Millican, J. Hoffmann, F. Song, J. Aslanides, S. Henderson,
R. Ring, S. Young, et al., Scaling language models: Methods, analysis & insights from training
gopher, arXiv preprint arXiv:2112.11446.

[104] M. Prince, Does active learning work? a review of the research, Journal of engineering education
93 (3) (2004) 223–231.

[105] A. Scheer, C. Noweski, C. Meinel, Transforming constructivist learning into action: Design thinking
in education, Design and Technology Education: An International Journal 17 (3).

[106] M. Forehand, Bloom’s taxonomy, Emerging perspectives on learning, teaching, and technology 41 (4)
(2010) 47–56.

[107] D. L. McCabe, L. K. Trevino, K. D. Butterfield, Cheating in academic institutions: A decade of

research, Ethics & Behavior 11 (3) (2001) 219–232. doi:10.1207/s15327019eb1103_2.
URL https://doi.org/10.1207/s15327019eb1103_2

[108] I. Ajzen, The theory of planned behavior, Organizational behavior and human decision processes
50 (2) (1991) 179–211.

[109] M. L. Nistal, An experience of continuous assessment in telecommunication technologies engineering:

New costs for the teacher, IEEE Revista Iberoamericana de Tecnologias del Aprendizaje 8 (2) (2013)
90–95. doi:10.1109/rita.2013.2258225.
URL https://doi.org/10.1109/rita.2013.2258225

[110] B. Noonan, C. R. Duncan, Peer and self-assessment in high schools, Practical Assessment, Research,
and Evaluation 10 (1) (2005) 17.

[111] K. Boustani, A. C. Tally, Y. R. Kim, C. Nippert-Eng, Gaming the name: Player strategies for
adapting to name constraints in online videogames, in: Proceedings of the Annual Symposium on
Computer-Human Interaction in Play, CHI PLAY ’20, Association for Computing Machinery, New
York, NY, USA, 2020, p. 120–131. doi:10.1145/3410404.3414259.
URL https://doi.org/10.1145/3410404.3414259

[112] A. S. M. Noman, S. Das, S. Patil, Techies against facebook: Understanding negative sentiment
toward facebook via user generated content, in: Proceedings of the 2019 CHI Conference on Human
Factors in Computing Systems, CHI ’19, Association for Computing Machinery, New York, NY,

71
 USA, 2019, p. 1–15. doi:10.1145/3290605.3300698.
URL https://doi.org/10.1145/3290605.3300698

[113] C. Huang, W. Jiang, J. Wu, G. Wang, Personalized review recommendation based on users’ aspect
sentiment, ACM Trans. Internet Technol. 20 (4). doi:10.1145/3414841.
URL https://doi.org/10.1145/3414841

[114] P. Zhou, Unleasing chatgpt on the metaverse: Savior or destroyer?, arXiv preprint arXiv:2303.13856.

72