Abstract
This
work-in-progress
paper
explores
faculty
perspectives
regarding
student
use
of
Generative
Artificial
Intelligence
(GAI)
assistance
tools,
such
as
ChatGPT,
to
complete
engineering
coursework.
A
common
debate
in
and
computer
science
exists
about
how
should
address
GAI
tools
(i.e.,
prevent
their
usage
order
maintain
academic
integrity,
teach
students
the
new
technologies,
or
establish
regulatory
guidelines
higher
education).
While
continues
disrupt
traditional
educational
paradigms,
its
full
impacts
on
teaching
learning
are
currently
unknown.
Such
work
is
especially
useful
for
fields
science,
whose
lies
at
forefront
technological
advancement
more
readily
adopt
technologies
into
course
tasks.
discusses
preliminary
findings
an
intrinsic
qualitative
case
study
that
answers
research
question:
How
do
perceive
undergraduate
completion?
Data
were
collected
using
semi-structured
interviews
with
faculty,
including
civil,
mechanical,
electrical,
biological
science.
As
a
result,
this
lays
groundwork
extensive
domain
contributes
broader
discourse
role
emerging
shaping
future
education.
ChatGPT
represents
a
groundbreaking
AI
application
that
has
garnered
significant
attention
since
its
inception.
However,
despite
promising
potential,
ethical
implications
have
sparked
considerable
debate.
This
study
aims
to
examine
the
key
concerns
surrounding
governance
of
by
conducting
bibliometric
analysis
and
cluster-based
content
relevant
scientific
literature.
The
identifies
influential
authors,
countries,
pivotal
publications,
revealing
three
primary
categories
issues
associated
with
ChatGPT:
human-related
ethics,
academic
integrity
technical
literacy,
artificial
intelligence
(AI)
technology
ethics
derived
concerns.
Additionally,
further
refines
these
synthesizing
frequently
occurring
keywords.
Building
on
this
framework,
provides
comprehensive
discussion
major
challenges
faced
ChatGPT,
as
well
outlining
future
research
priorities.
Furthermore,
investigates
knowledge
base
underlying
ChatGPT's
governance,
exploring
high-citation
high-link-strength
literature
through
co-citation
analysis,
thereby
mapping
landscape
highlighting
areas
growing
scholarly
interest.
offers
valuable
insights
for
policymakers,
researchers,
practitioners,
emphasizing
need
more
stringent
policies,
guidelines,
robust
design
in
development
similar
technologies.
iScience,
Год журнала:
2024,
Номер
27(4), С. 109334 - 109334
Опубликована: Фев. 24, 2024
With
advancements
in
large
language
models,
artificial
intelligence
(AI)
is
undergoing
a
paradigm
shift
where
AI
models
can
be
repurposed
with
minimal
effort
across
various
downstream
tasks.
This
provides
great
promise
learning
generally
useful
representations
from
biomedical
corpora,
at
scale,
which
would
empower
solutions
healthcare
and
research.
Nonetheless,
our
understanding
of
how
they
work,
when
fail,
what
are
capable
remains
underexplored
due
to
their
emergent
properties.
Consequently,
there
need
comprehensively
examine
the
use
biomedicine.
review
aims
summarize
existing
studies
biomedicine
identify
topics
ripe
for
future
research,
along
technical
analytical
challenges
w.r.t.
interpretability.
We
expect
this
help
researchers
practitioners
better
understand
landscape
methods
available
enhance
interpretability
models.
Artificial Intelligence Review,
Год журнала:
2024,
Номер
57(10)
Опубликована: Авг. 16, 2024
The
fast
evolution
of
artificial
intelligence
frameworks
has
resulted
in
the
creation
increasingly
sophisticated
large
language
models
(LLM),
ChatGPT
being
most
famous
one.
This
study
paper
dives
into
this
LLM
with
a
case
ChatGPT's
architecture
and
provides
thorough
comparative
analysis
its
numerous
versions,
tracking
history
from
conception
to
recent
incarnations.
research
intends
give
full
knowledge
model's
by
investigating
underlying
mechanisms
enhancements
provided
each
edition.
covers
key
aspects
such
as
model
size,
training
data,
fine-tuning
techniques,
performance
metrics.
Furthermore,
evaluates
limits
many
These
limitations
include
common
sense
reasoning
difficulties,
biased
replies,
verbosity,
sensitivity
input
wording,
others.
Each
constraint
is
investigated
for
potential
remedies
workarounds.
article
also
complete
progress
through
multiple
iterations.
It
gives
vital
insights
academics,
developers,
users
wanting
harness
promise
while
managing
restrictions
exploring
both
strengths
limitations.
distinctiveness
rests
comprehensive
assessment
architectural
development
practical
strategy
resolving
myriad
difficulties
producing
cohesive
contextually
relevant
replies.
Electronics,
Год журнала:
2024,
Номер
13(24), С. 5043 - 5043
Опубликована: Дек. 22, 2024
In
the
era
of
large
AI
models,
intricate
architectures
and
vast
parameter
sets
models
such
as
language
(LLMs)
present
significant
challenges
for
effective
quality
management
(AIQM).
This
paper
investigates
assurance
a
specific
LLM-based
product:
ChatGPT-based
sentiment
analysis.
The
study
focuses
on
stability
issues,
examining
both
operation
robustness
ChatGPT’s
underlying
large-scale
model.
Through
experimental
analysis
benchmark
datasets
analysis,
findings
highlight
analysis’s
susceptibility
to
uncertainty,
which
relates
various
operational
factors.
Furthermore,
reveals
that
model
faces
challenges,
particularly
when
confronted
with
conventional
small-text
adversarial
attacks
targeting
robustness.
Oeconomia Copernicana,
Год журнала:
2025,
Номер
2025(16), С. 79 - 124
Опубликована: Март 30, 2025
Research
background:
Social
media
plays
a
crucial
role
today
in
enhancing
or
limiting
how
fake
news
is
spread.
Whether
devised
by
man
developed
artificial
intelligence,
it
has
the
power
to
rapidly
change
consumers’
minds,
encouraging
them
adopt
new
behaviors,
perceive
situations
differently,
even
act
total
opposition
what
might
be
expected.
The
dynamics
of
communication
highlights
need
for
an
organizational
response
adapted
AI
technologies
and
dissemination
within
social
networks.
Purpose
this
article:
This
paper
aims
reveal,
means
bibliometric
analysis
systematic
literature
review,
generative
capabilities
intelligence
creation
spread
business
environment,
acknowledging
previous
research
predicting
accurately
constant
developments
contemporary
society.
Methods:
based
on
PRISMA
flowchart
examine
contribute
whilst
also
highlighting
potential
regulations
standards
false
information.
Initially,
database
included
over
3,400
highly
cited
articles
retrieved
from
Scopus
Web
Science,
published
last
years,
which
203
were
selected
inclusion
analysis.
follows
directions
related
detection
methods
strategies,
legislation
policies
governing
used
connected
environment.
Fake
typologies
relating
advancement
are
explored.
Findings
&
value
added:
By
analysing
important
phrases,
including
information,
misinformation,
disinformation,
mal-information,
deepfakes,
investigates
categorization
linked
environment
concepts.
It
underscores
better
truth
comprehension
significance
fact-checking
preventing
with
governance
institutional
implications
terms
economics
intelligence-generated
market.
While
studies
have
examined
phenomenon
several
angles,
there
still
gap,
as
concentrates
more
consumed
rather
than
created.
bridge
gap
providing
comprehensive
examination
perspectives
typology,
creation,
detection,
regulatory
means.
When
used
in
conjunction
with
education,
ChatGPT
is
a
helpful
tool
for
teaching
and
understanding
fundamental
concepts.
Students
who
struggle
to
find
study
partners
or
lack
the
time
attend
tutoring
sessions
may
discover
that
reviewing
studying
chatGPT
are
helpful.
Students'
access
material
completion
of
assignments
can
be
accelerated
using
this
technology
research
education.
More
than
one-third
university
students
participated
poll
said
they
writing
assessments.
Professors
have
referred
these
technologies
as
"threat"
"plague
on
education,"
has
been
prohibited
from
several
educational
institutions.
Thus,
focuses
introduction
Features,
Challenges,
Threads
Higher
Education.
This
conducted
literature
review
method.
According
result,
boost
student
engagement
by
providing
immersive,
dynamic,
customized
learning
environments.
However,
there
some
related
issues
need
considered
such
integrity,
accuracy
reliability,
information
bias,
privacy
issues.
We
therefore
infer
benefits
well
drawbacks
employing
AI
education
taken
into
account.
AI and Ethics,
Год журнала:
2024,
Номер
5(1), С. 653 - 675
Опубликована: Янв. 23, 2024
ChatGPT,
a
chatbot
using
the
GPT-n
series
large
language
model,
has
surged
in
popularity
by
providing
conversation,
assistance,
and
entertainment.
This
raised
questions
about
its
agency
resulting
implications
on
trust
blame,
particularly
when
concerning
portrayal
social
media
platforms
like
Twitter.
Understanding
blame
is
crucial
for
gauging
public
perception,
reliance
on,
adoption
of
AI-driven
tools
ChatGPT.
To
explore
ChatGPT's
perceived
status
as
an
algorithmic
actor
uncover
through
transitivity,
we
examined
88,058
tweets
published
'hype
period'
between
November
2022
March
2023,
Corpus
Linguistics
Critical
Discourse
Analysis,
underpinned
Social
Actor
Representation.
Notably,
ChatGPT
was
presented
87%
occasions,
personalisation
metaphor
to
emphasise
role
content
creation,
information
dissemination,
influence.
However,
dynamic
presentation,
oscillating
creative
source,
reflected
users'
uncertainty
regarding
capabilities
and,
thus,
attribution
occurred.
On
13%
passively
backgrounding
exclusion.
Here,
emphasis
informing
influencing
underscores
interactors'
it
information,
bearing
dissemination
AI-generated
content.
Therefore,
this
study
contributes
understanding
decision-making
algorithms
their
valuable
AI
developers
policymakers
relevant
comprehending
dealing
with
power
dynamics
today's
age
AI.
Frontiers in Artificial Intelligence,
Год журнала:
2024,
Номер
6
Опубликована: Янв. 4, 2024
The
latest
developments
of
chatbots
driven
by
Large
Language
Models
(LLMs),
more
specifically
ChatGPT,
have
shaken
the
foundations
how
text
is
created,
and
may
drastically
reduce
change
need,
ability,
valuation
human
writing.
Furthermore,
our
trust
in
written
word
likely
to
decrease,
as
an
increasing
proportion
all
will
be
AI-generated
–
potentially
incorrect.
In
this
essay,
I
discuss
these
implications
possible
scenarios
for
us
humans,
AI
itself.
