arXiv (Cornell University),
Год журнала:
2023,
Номер
unknown
Опубликована: Янв. 1, 2023
Chat
Generative
Pre-trained
Transformer
(ChatGPT)
has
gained
significant
interest
and
attention
since
its
launch
in
November
2022.
It
shown
impressive
performance
various
domains,
including
passing
exams
creative
writing.
However,
challenges
concerns
related
to
biases
trust
persist.
In
this
work,
we
present
a
comprehensive
review
of
over
100
Scopus-indexed
publications
on
ChatGPT,
aiming
provide
taxonomy
ChatGPT
research
explore
applications.
We
critically
analyze
the
existing
literature,
identifying
common
approaches
employed
studies.
Additionally,
investigate
diverse
application
areas
where
found
utility,
such
as
healthcare,
marketing
financial
services,
software
engineering,
academic
scientific
writing,
education,
environmental
science,
natural
language
processing.
Through
examining
these
applications,
gain
valuable
insights
into
potential
addressing
real-world
challenges.
also
discuss
crucial
issues
trustworthiness,
emphasizing
need
for
further
development
areas.
Furthermore,
identify
future
directions
research,
proposing
solutions
current
speculating
expected
advancements.
By
fully
leveraging
capabilities
can
unlock
across
leading
advancements
conversational
AI
transformative
impacts
society.
Journal of King Saud University - Computer and Information Sciences,
Год журнала:
2023,
Номер
35(8), С. 101675 - 101675
Опубликована: Авг. 2, 2023
Chat
Generative
Pre-trained
Transformer
(ChatGPT)
has
gained
significant
interest
and
attention
since
its
launch
in
November
2022.
It
shown
impressive
performance
various
domains,
including
passing
exams
creative
writing.
However,
challenges
concerns
related
to
biases
trust
persist.
In
this
work,
we
present
a
comprehensive
review
of
over
100
Scopus-indexed
publications
on
ChatGPT,
aiming
provide
taxonomy
ChatGPT
research
explore
applications.
We
critically
analyze
the
existing
literature,
identifying
common
approaches
employed
studies.
Additionally,
investigate
diverse
application
areas
where
found
utility,
such
as
healthcare,
marketing
financial
services,
software
engineering,
academic
scientific
writing,
education,
environmental
science,
natural
language
processing.
Through
examining
these
applications,
gain
valuable
insights
into
potential
addressing
real-world
challenges.
also
discuss
crucial
issues
trustworthiness,
emphasizing
need
for
further
development
areas.
Furthermore,
identify
future
directions
research,
proposing
solutions
current
speculating
expected
advancements.
By
fully
leveraging
capabilities
can
unlock
across
leading
advancements
conversational
AI
transformative
impacts
society.
JMIR Medical Education,
Год журнала:
2024,
Номер
10, С. e50965 - e50965
Опубликована: Фев. 8, 2024
Background
The
potential
of
artificial
intelligence
(AI)–based
large
language
models,
such
as
ChatGPT,
has
gained
significant
attention
in
the
medical
field.
This
enthusiasm
is
driven
not
only
by
recent
breakthroughs
and
improved
accessibility,
but
also
prospect
democratizing
knowledge
promoting
equitable
health
care.
However,
performance
ChatGPT
substantially
influenced
input
language,
given
growing
public
trust
this
AI
tool
compared
to
that
traditional
sources
information,
investigating
its
accuracy
across
different
languages
particular
importance.
Objective
study
aimed
compare
GPT-3.5
GPT-4
with
students
on
written
German
licensing
examination.
Methods
To
assess
GPT-3.5’s
GPT-4's
proficiency,
we
used
937
original
multiple-choice
questions
from
3
examinations
October
2021,
April
2022,
2022.
Results
achieved
an
average
score
85%
ranked
92.8th,
99.5th,
92.6th
percentiles
among
who
took
same
respectively.
represents
a
substantial
improvement
27%
GPT-3.5,
which
passed
1
out
examinations.
While
performed
well
psychiatry
questions,
exhibited
strengths
internal
medicine
surgery
showed
weakness
academic
research.
Conclusions
results
highlight
ChatGPT’s
remarkable
moderate
(GPT-3.5)
high
competency
(GPT-4)
answering
examination
German.
GPT-4’s
predecessor
was
imprecise
inconsistent,
it
demonstrates
considerable
improve
education
patient
care,
provided
medically
trained
users
critically
evaluate
results.
As
replacement
search
engines
tools
seems
possible
future,
further
studies
nonprofessional
are
needed
safety
for
general
population.
Education Sciences,
Год журнала:
2024,
Номер
14(8), С. 814 - 814
Опубликована: Июль 25, 2024
This
paper
investigates
the
integration
of
ChatGPT
into
educational
environments,
focusing
on
its
potential
to
enhance
personalized
learning
and
ethical
concerns
it
raises.
Through
a
systematic
literature
review,
interest
analysis,
case
studies,
research
scrutinizes
application
in
diverse
contexts,
evaluating
impact
teaching
practices.
The
key
findings
reveal
that
can
significantly
enrich
education
by
offering
dynamic,
experiences
real-time
feedback,
thereby
boosting
efficiency
learner
engagement.
However,
study
also
highlights
significant
challenges,
such
as
biases
AI
algorithms
may
distort
content
inability
replicate
emotional
interpersonal
dynamics
traditional
teacher–student
interactions.
acknowledges
fast-paced
evolution
technologies,
which
render
some
obsolete,
underscoring
need
for
ongoing
adapt
strategies
accordingly.
provides
balanced
analysis
opportunities
challenges
education,
emphasizing
considerations
strategic
insights
responsible
technologies.
These
are
valuable
educators,
policymakers,
researchers
involved
digital
transformation
education.
PLOS Digital Health,
Год журнала:
2024,
Номер
3(5), С. e0000492 - e0000492
Опубликована: Май 2, 2024
Background
The
rapid
evolution
of
conversational
and
generative
artificial
intelligence
(AI)
has
led
to
the
increased
deployment
AI
tools
in
healthcare
settings.
While
these
promise
efficiency
expanded
access
services,
there
are
growing
concerns
ethically,
practically
terms
inclusivity.
This
study
aimed
identify
activities
which
reduce
bias
make
their
designs
implementation
more
equitable.
Methods
A
qualitative
research
approach
was
employed
develop
an
analytical
framework
based
on
content
analysis
17
guidelines
about
use
clinical
stakeholder
consultation
subsequently
conducted
with
a
total
33
ethnically
diverse
community
members,
designers,
industry
experts
relevant
health
professionals
further
roadmap
for
equitable
design
healthcare.
Framework
interview
data.
Results
10-stage
developed
outline
phases:
1)
Conception
planning,
2)
Diversity
collaboration,
3)
Preliminary
research,
4)
Co-production,
5)
Safety
measures,
6)
testing,
7)
Healthcare
integration,
8)
Service
evaluation
auditing,
9)
Maintenance,
10)
Termination.
Discussion
We
have
made
specific
recommendations
increase
AI’s
equity
as
part
services.
These
emphasise
importance
collaborative
involvement
patient
groups
navigating
technologies.
Further
must
assess
impact
recommended
chatbots’
fairness
ability
inequalities.
ChatGPT,
a
new
language
model
developed
by
OpenAI,
has
garnered
significant
attention
in
various
fields
since
its
release.
This
literature
review
provides
an
overview
of
early
ChatGPT
across
multiple
disciplines,
exploring
applications,
limitations,
and
ethical
considerations.
The
encompasses
Scopus-indexed
publications
from
November
2022
to
April
2023
includes
156
articles
related
ChatGPT.
findings
reveal
predominance
negative
sentiment
though
subject-specific
attitudes
must
be
considered.
highlights
the
implications
many
including
healthcare,
raising
concerns
about
employment
opportunities
While
holds
promise
for
improved
communication,
further
research
is
needed
address
capabilities
limitations.
insights
into
on
informing
future
investigations
practical
applications
chatbot
technology,
as
well
development
usage
generative
AI.
Sensors,
Год журнала:
2023,
Номер
23(10), С. 4879 - 4879
Опубликована: Май 18, 2023
In
this
paper,
we
investigated
the
applicability
of
ChatGPT
AI
in
electronics
research
and
development
via
a
case
study
applied
sensors
embedded
electronic
systems,
topic
that
is
rarely
mentioned
recent
literature,
thus
providing
new
insight
for
professionals
academics.
The
initial
electronics-development
tasks
smart
home
project
were
prompted
to
system
find
out
its
capabilities
limitations.
We
wanted
obtain
detailed
information
on
central
processing
controller
units
actual
usable
specific
project,
their
specifications
recommendations
hardware
software
design
flow
additionally.
Furthermore,
an
extensive
literature
survey
was
requested
see
if
bot
could
offer
scientific
papers
covering
given
topic.
It
found
responded
with
proper
controllers.
However,
suggested
sensor
units,
only
partially
acceptable,
occasional
errors
generated
code.
results
showed
non-acceptable,
fabricated
citations
(fake
authors
list,
title,
journal
details
DOI-Digital
Object
identifier)
presented
by
bot.
paper
provides
qualitative
analysis,
performance
analysis
critical
discussion
aforementioned
aspects
while
query
set,
answers
codes
as
supplied
data
goal
give
added
value
researchers
developers
trying
reach
tools
profession.
VFAST Transactions on Software Engineering,
Год журнала:
2024,
Номер
12(1), С. 152 - 168
Опубликована: Март 31, 2024
Artificial
intelligence
(AI)
is
the
main
branch
of
computer
science
that
permits
advanced
machines
to
interpret
and
analyze
complex
healthcare
data
elaborating
recent
challenges
in
medical
field
study.
The
current
state
AI
applications
examined
this
systematic
literature
review,
with
an
emphasis
on
technology's
accomplishments,
difficulties,
potential.
wide
breadth
technologies
used
settings,
such
as
robots,
vision,
machine
learning,
natural
language
processing,
highlighted
review
through
extensive
analysis
peer-reviewed
publications.
It
talks
about
how
customized
medicine,
predictive
analytics,
illness
detection,
treatment
planning
are
just
a
few
areas
delivery
AI-driven
transforming.
According
research
by
investment
bank
Goldman
Sachs,
300
million
full-time
employees
could
be
replaced
artificial
(AI).
In
US
Europe,
it
might
replace
25%
labor
duties,
but
also
lead
increase
productivity
creation
new
jobs.
Additionally,
eventually
result
7%
rise
global
annual
value
products
services
produced.
paper
projects
approximately
all
employment
performed
totally
two-thirds
jobs
U.S.
Europe
"are
exposed
some
degree
automation.
"The
most
likely
groups
impacted
workforce
automation
educated
white-collar
workers
making
up
$80,000
annually,
according
from
OpenAI
University
Pennsylvania.
McKinsey
Global
Institute
study,
developments
digitalization,
may
require
at
least
14%
worldwide
change
2030.
