Journal of Physics Conference Series,
Journal Year:
2024,
Volume and Issue:
2871(1), P. 012023 - 012023
Published: Oct. 1, 2024
Abstract
The
article,
based
on
empirical
and
theoretical
research,
reveals
the
phenomenology
of
transformations
human
cognitive
sphere
when
interacting
with
artificial
intelligence.
analysis
indicated
changes
in
is
carried
out
basis
“Concept
multi-channel
Human-Computer
interaction”
developed
by
us.
essence
this
concept
that
interaction
intelligence
implemented
actualization
formation
typical
phenomena.
These
phenomena
are
considered
systemically
multifunctionally,
namely
as
relatively
independent
cognitive:
types
interactions,
stages,
strategies,
channels,
ontologies.
Within
conceptual
substantive
framework
concept,
we
distinguish
following
cognition
(channels,
etc.):
I
–
orientational-cognitive;
II
subject-cognitive;
III
communicative
cognitive;
IV
analytical;
V
hermeneutic;
VI-cognitive-ontological;
VII
creative.
identification
interactions
aimed
at
its
representation
a
complex,
dynamic,
multidimensional,
multichannel
intellectual
system,
features
which
significant
for
educational
sociocultural
practices,
well
further
development
technologies,
including
functional
orientation
specificity,
ergonomics,
architecture,
design
interface.
A
study
was
conducted
among
students
higher
education
institutions
determining
specificity
(structure)
“Human
Artificial
Intelligence”
system.
Based
results
distribution
answers
each
test
questions
interpretation
cluster
(the
Canopy
algorithm
used),
dominance
“I
orientational-cognitive”
type
determined,
indicates
rather
but
initial
interest
technologies.
There
also
even
all
other
interactions.
above
novelty
innovation
technology.
This
correlates
respondents
having
different
cognition,
namely:
orientational,
analytical-synthetic,
conceptual,
interpretive,
ontological,
creative
thinking,
corresponding
intentions
motivation
to
use
tools
various
spheres
activity.
Information,
Journal Year:
2025,
Volume and Issue:
16(2), P. 117 - 117
Published: Feb. 7, 2025
The
emergence
of
generative
artificial
intelligence
(GAI)
has
revolutionized
numerous
aspects
our
lives
and
presents
significant
opportunities
in
education.
However,
specific
digital
competencies
are
essential
to
effectively
leverage
this
technology’s
potential.
Notably,
prompt
engineering
proficiency
a
barrier
achieving
optimal
outcomes.
In
response,
various
solutions
being
developed,
including
custom
GPTs
available
through
OpenAI’s
ChatGPT
platform.
This
study
validates
‘GamifIcA
Edu’,
specialized
GPT-based
assistant
for
gamification
serious
games,
designed
enable
educators
implement
these
pedagogical
approaches
without
requiring
advanced
expertise.
is
achieved
the
utilization
pre-designed
instructional
frameworks.
assistant’s
effectiveness
was
evaluated
using
comprehensive
rubric
across
five
distinct
use-case
scenarios.
Each
scenario
underwent
four
different
tests,
representing
varied
learning
contexts
multiple
academic
disciplines.
validation
methodology
involved
systematic
assessment
performance
diverse
educational
settings.
findings
demonstrate
successful
implementation
custom-designed
GPT,
which
generated
contextually
appropriate
responses
natural
language
interactions,
thus
eliminating
need
complex
structures.
research
highlights
potential
instruction-based
design
development
AI
assistants
that
empower
users
with
limited
knowledge
achieve
expert-level
results.
These
have
implications
democratization
AI-enhanced
tools.
BioMedInformatics,
Journal Year:
2025,
Volume and Issue:
5(1), P. 12 - 12
Published: Feb. 27, 2025
Background:
Large
Language
Models
(LLMs)
have
demonstrated
strong
performances
in
clinical
question-answering
(QA)
benchmarks,
yet
their
effectiveness
addressing
real-world
consumer
medical
queries
remains
underexplored.
This
study
evaluates
the
capabilities
and
limitations
of
LLMs
answering
health
questions
using
MedRedQA
dataset,
which
consists
answers
by
verified
experts
from
AskDocs
subreddit.
Methods:
Five
LLMs-GPT-4o
mini,
Llama
3.1-70B,
Mistral-123B,
Mistral-7B,
Gemini-Flash
were
assessed
a
cross-evaluation
framework.
Each
model
generated
responses
to
outputs
evaluated
every
comparing
them
with
expert
responses.
Human
evaluation
was
used
assess
reliability
models
as
evaluators.
Results:
GPT-4o
mini
achieved
highest
alignment
according
four
out
five
models’
judges,
while
Mistral-7B
scored
lowest
three
judges.
Overall,
show
low
Conclusions:
Current
small
or
medium
sized
struggle
provide
accurate
must
be
significantly
improved.
Information,
Journal Year:
2025,
Volume and Issue:
16(4), P. 264 - 264
Published: March 26, 2025
This
paper
explores
the
potential
of
AI-based
digital
tutors
to
enhance
student
learning
by
providing
accurate,
course-specific
answers
complex
questions,
anchored
in
validated
course
materials.
The
Tel
Aviv
University
Digital
Tutor
(TAUDT)
exemplifies
this
approach,
enabling
students
navigate
and
comprehend
academic
content
with
ease.
By
citing
specific
passages
materials,
TAUDT
ensures
pedagogical
accuracy
relevance
while
fostering
independent
learning.
Its
modular
design
allows
for
seamless
integration
advancements
AI
state-of-the-art
technologies,
ensuring
long-term
adaptability
performance.
Designed
integrate
effortlessly
into
existing
workflows,
requires
no
technological
expertise
from
instructors,
addressing
barriers
posed
technophobia
among
faculty.
A
pilot
study
demonstrated
high
levels
engagement,
highlighting
its
as
a
scalable,
adaptive
solution
higher
education.
Research Square (Research Square),
Journal Year:
2025,
Volume and Issue:
unknown
Published: March 31, 2025
Abstract
The
integration
of
Large
Language
Models
(LLMs)
with
Learning
Management
Systems
(LMSs)
has
the
potential
to
enhance
task
automation
and
accessibility
in
education.
However,
hallucination
where
LLMs
generate
inaccurate
or
misleading
information
remains
a
significant
challenge.
This
study
introduces
Dynamic
Course
Content
Integration
(DCCI)
mechanism,
which
dynamically
retrieves
integrates
course
content
curriculum
from
Canvas
LMS
into
LLM-powered
assistant,
Ask
ME.
By
employing
prompt
engineering
structure
retrieved
within
LLM’s
context
window,
DCCI
ensures
accuracy,
relevance,
contextual
alignment,
mitigating
hallucination.
To
evaluate
DCCI’s
effectiveness,
ME’s
usability,
broader
student
perceptions
AI
education,
mixed-methods
approach
was
employed,
incorporating
user
satisfaction
ratings
structured
survey.
Results
pilot
indicate
high
(4.614/5),
students
recognizing
ability
provide
timely
contextually
relevant
responses
for
both
administrative
course-related
inquiries.
Additionally,
majority
agreed
that
reduced
platform-switching,
improving
engagement,
comprehension.
AI’s
role
reducing
classroom
hesitation
fostering
self-directed
learning
intellectual
curiosity
also
highlighted.
Despite
these
benefits
positive
perception
tools,
concerns
emerged
regarding
over-reliance
on
AI,
accuracy
limitations,
ethical
issues
such
as
plagiarism
student-teacher
interaction.
These
findings
emphasize
need
strategic
implementation,
safeguards,
pedagogical
framework
prioritizes
human-AI
collaboration
over
substitution.
contributes
AI-enhanced
education
by
demonstrating
how
context-aware
retrieval
mechanisms
like
improve
LLM
reliability
educational
engagement
while
ensuring
responsible
integration.
