Internet of Things and Cyber-Physical Systems,
Год журнала:
2023,
Номер
3, С. 121 - 154
Опубликована: Янв. 1, 2023
In
recent
years,
artificial
intelligence
(AI)
and
machine
learning
have
been
transforming
the
landscape
of
scientific
research.
Out
which,
chatbot
technology
has
experienced
tremendous
advancements
in
especially
with
ChatGPT
emerging
as
a
notable
AI
language
model.
This
comprehensive
review
delves
into
background,
applications,
key
challenges,
future
directions
ChatGPT.
We
begin
by
exploring
its
origins,
development,
underlying
technology,
before
examining
wide-ranging
applications
across
industries
such
customer
service,
healthcare,
education.
also
highlight
critical
challenges
that
faces,
including
ethical
concerns,
data
biases,
safety
issues,
while
discussing
potential
mitigation
strategies.
Finally,
we
envision
areas
further
research
focusing
on
integration
other
technologies,
improved
human-AI
interaction,
addressing
digital
divide.
offers
valuable
insights
for
researchers,
developers,
stakeholders
interested
ever-evolving
AI-driven
conversational
agents.
study
explores
various
ways
revolutionizing
research,
spanning
from
processing
hypothesis
generation
to
collaboration
public
outreach.
Furthermore,
paper
examines
concerns
surrounding
use
highlighting
importance
striking
balance
between
AI-assisted
innovation
human
expertise.
The
presents
several
issues
existing
computing
domain
how
can
invoke
notion.
work
includes
some
biases
limitations
It
is
worth
note
despite
controversies
attracted
remarkable
attentions
academia,
very
short
span
time.
Applied Sciences,
Год журнала:
2023,
Номер
13(9), С. 5783 - 5783
Опубликована: Май 8, 2023
In
recent
years,
the
rise
of
advanced
artificial
intelligence
technologies
has
had
a
profound
impact
on
many
fields,
including
education
and
research.
One
such
technology
is
ChatGPT,
powerful
large
language
model
developed
by
OpenAI.
This
offers
exciting
opportunities
for
students
educators,
personalized
feedback,
increased
accessibility,
interactive
conversations,
lesson
preparation,
evaluation,
new
ways
to
teach
complex
concepts.
However,
ChatGPT
poses
different
threats
traditional
research
system,
possibility
cheating
online
exams,
human-like
text
generation,
diminished
critical
thinking
skills,
difficulties
in
evaluating
information
generated
ChatGPT.
study
explores
potential
that
overall
from
perspective
educators.
Furthermore,
programming
learning,
we
explore
how
helps
improve
their
skills.
To
demonstrate
this,
conducted
coding-related
experiments
with
code
generation
problem
descriptions,
pseudocode
algorithms
texts,
correction.
The
codes
are
validated
an
judge
system
evaluate
accuracy.
addition,
several
surveys
teachers
find
out
supports
learning
teaching.
Finally,
present
survey
results
analysis.
Natural Language Processing Journal,
Год журнала:
2023,
Номер
6, С. 100048 - 100048
Опубликована: Дек. 19, 2023
Large
language
models
(LLMs)
are
a
special
class
of
pretrained
(PLMs)
obtained
by
scaling
model
size,
pretraining
corpus
and
computation.
LLMs,
because
their
large
size
on
volumes
text
data,
exhibit
abilities
which
allow
them
to
achieve
remarkable
performances
without
any
task-specific
training
in
many
the
natural
processing
tasks.
The
era
LLMs
started
with
OpenAI's
GPT-3
model,
popularity
has
increased
exponentially
after
introduction
like
ChatGPT
GPT4.
We
refer
its
successor
OpenAI
models,
including
GPT4,
as
family
(GLLMs).
With
ever-rising
GLLMs,
especially
research
community,
there
is
strong
need
for
comprehensive
survey
summarizes
recent
progress
multiple
dimensions
can
guide
community
insightful
future
directions.
start
paper
foundation
concepts
transformers,
transfer
learning,
self-supervised
models.
then
present
brief
overview
GLLMs
discuss
various
downstream
tasks,
specific
domains
languages.
also
data
labelling
augmentation
robustness
effectiveness
evaluators,
finally,
conclude
To
summarize,
this
will
serve
good
resource
both
academic
industry
people
stay
updated
latest
related
GLLMs.
This
paper
provides
a
survey
of
the
emerging
area
Large
Language
Models
(LLMs)
for
Software
Engineering
(SE).
It
also
sets
out
open
research
challenges
application
LLMs
to
technical
problems
faced
by
software
engineers.
LLMs'
emergent
properties
bring
novelty
and
creativity
with
applications
right
across
spectrum
activities
including
coding,
design,
requirements,
repair,
refactoring,
performance
improvement,
documentation
analytics.
However,
these
very
same
pose
significant
challenges;
we
need
techniques
that
can
reliably
weed
incorrect
solutions,
such
as
hallucinations.
Our
reveals
pivotal
role
hybrid
(traditional
SE
plus
LLMs)
have
play
in
development
deployment
reliable,
efficient
effective
LLM-based
SE.
Frontiers in Artificial Intelligence,
Год журнала:
2023,
Номер
6
Опубликована: Окт. 31, 2023
The
release
of
ChatGPT
has
initiated
new
thinking
about
AI-based
Chatbot
and
its
application
drawn
huge
public
attention
worldwide.
Researchers
doctors
have
started
the
promise
AI-related
large
language
models
in
medicine
during
past
few
months.
Here,
comprehensive
review
highlighted
overview
their
current
role
medicine.
Firstly,
general
idea
Chatbots,
evolution,
architecture,
medical
use
are
discussed.
Secondly,
is
discussed
with
special
emphasis
medicine,
architecture
training
methods,
diagnosis
treatment,
research
ethical
issues,
a
comparison
other
NLP
illustrated.
article
also
limitations
prospects
ChatGPT.
In
future,
these
will
immense
healthcare.
However,
more
needed
this
direction.
Computers and Education Artificial Intelligence,
Год журнала:
2024,
Номер
6, С. 100219 - 100219
Опубликована: Апрель 3, 2024
This
study
investigates
the
impact
of
activity-based
learning
and
utilization
ChatGPT
on
students'
academic
performance
within
educational
framework.
The
aims
to
assess
effectiveness
in
comparison
traditional
methods,
while
also
evaluating
potential
benefits
drawbacks
integrating
as
an
tool.
employs
a
comparative
approach,
analyzing
outcomes
students
exposed
versus
those
using
conventional
methods.
Additionally,
examines
usage
education
through
surveys
trials
determine
its
contribution
personalized
feedback,
interactive
learning,
innovative
teaching
findings
reveal
that
enhances
engagement,
motivation,
critical
thinking
skills.
Students
participating
demonstrate
improved
achievement,
which
is
attributed
their
active
involvement
practical
application
knowledge.
Similarly,
integration
offers
novel
avenues
for
individualized
assistance,
fostering
understanding
exploration
complex
concepts.
In
conclusion,
proves
be
student-centered
approach
by
participation
engagement.
showcases
enhance
experiences
conversations
methodologies,
despite
considerations
regarding
limitations
ethical
implications.
Findings of the Association for Computational Linguistics: ACL 2022,
Год журнала:
2023,
Номер
unknown
Опубликована: Янв. 1, 2023
The
development
of
large
language
models
(LLMs)
such
as
ChatGPT
has
brought
a
lot
attention
recently.
However,
their
evaluation
in
the
benchmark
academic
datasets
remains
under-explored
due
to
difficulty
evaluating
generative
outputs
produced
by
this
model
against
ground
truth.
In
paper,
we
aim
present
thorough
ChatGPT's
performance
on
diverse
datasets,
covering
tasks
like
question-answering,
text
summarization,
code
generation,
commonsense
reasoning,
mathematical
problem-solving,
machine
translation,
bias
detection,
and
ethical
considerations.
Specifically,
evaluate
across
140
analyze
255K
responses
it
generates
these
datasets.
This
makes
our
work
largest
NLP
benchmarks.
short,
study
aims
validate
strengths
weaknesses
various
provide
insights
for
future
research
using
LLMs.
We
also
report
new
emergent
ability
follow
multi-query
instructions
that
mostly
found
other
instruction-tuned
models.
Our
extensive
shows
even
though
is
capable
performing
wide
variety
tasks,
may
obtain
impressive
several
still
far
from
achieving
reliably
solve
many
challenging
tasks.
By
providing
assessment
paper
sets
stage
targeted
deployment
ChatGPT-like
LLMs
real-world
applications.
Family Medicine and Community Health,
Год журнала:
2023,
Номер
11(4), С. e002391 - e002391
Опубликована: Сен. 1, 2023
To
compare
evaluations
of
depressive
episodes
and
suggested
treatment
protocols
generated
by
Chat
Generative
Pretrained
Transformer
(ChatGPT)-3
ChatGPT-4
with
the
recommendations
primary
care
physicians.Vignettes
were
input
to
ChatGPT
interface.
These
vignettes
focused
primarily
on
hypothetical
patients
symptoms
depression
during
initial
consultations.
The
creators
these
meticulously
designed
eight
distinct
versions
in
which
they
systematically
varied
patient
attributes
(sex,
socioeconomic
status
(blue
collar
worker
or
white
worker)
severity
(mild
severe)).
Each
variant
was
subsequently
introduced
into
ChatGPT-3.5
ChatGPT-4.
vignette
repeated
10
times
ensure
consistency
reliability
responses.For
mild
depression,
recommended
psychotherapy
95.0%
97.5%
cases,
respectively.
Primary
physicians,
however,
only
4.3%
cases.
For
severe
favoured
an
approach
that
combined
psychotherapy,
while
physicians
a
approach.
pharmacological
showed
preference
for
exclusive
use
antidepressants
(74%
68%,
respectively),
contrast
who
typically
mix
anxiolytics/hypnotics
(67.4%).
Unlike
no
gender
biases
its
recommendations.ChatGPT-3.5
aligned
well
accepted
guidelines
managing
without
showing
observed
among
physicians.
Despite
potential
benefit
using
atificial
intelligence
(AI)
chatbots
like
enhance
clinical
decision
making,
further
research
is
needed
refine
AI
cases
consider
risks
ethical
issues.
Humanities and Social Sciences Communications,
Год журнала:
2024,
Номер
11(1)
Опубликована: Фев. 26, 2024
Abstract
The
emergence
of
ChatGPT
has
sparked
new
expectations
for
AI-empowered
educational
transformation.
However,
it
remains
unknown
which
factors
affect
its
effectiveness
in
empowering
learners
to
solve
programming
problems.
Therefore,
this
study
employed
a
quasi-experimental
research
design
and
used
Python
graphing
education
as
an
example
investigate
the
influencing
applying
problem-solving.
Findings:
AI
literacy
significantly
influences
learners’
using
problems,
with
awareness
usage
being
key
factors.
knowledge
base
language
affects
Learners’
cognitive
level
their
problem-solving,
while
intention
does
not
have
significant
impact.
use
improves
after
application.
Based
on
these
findings,
proposes
that
process
Artificial
Intelligence
Generated
Content
(AIGC)
products,
focus
should
shift
from
cultivating
literacy,
laying
foundation
learning
AIGC
products.
It
is
suggested
mastering
specific
graph-based
rules
method
Additionally,
enhancing
technology
strengthen
technological
awareness,
thereby
creating
practical
pathways
