MJSAT Malaysian Journal of Science

and
Advanced Technology

journal homepage: https://mjsat.com.my/

Artificial Intelligence (AI) and Machine Learning Application in ChatGPT: A

Review

Anne Dashini Kannan1

1
Department of Computer Engineering and Computer Science, School of Engineering and Computing, Manipal
International University, Negeri Sembilan,Malaysia

KEYWORDS ABSTRACT

ChatGPT
Artificial Intelligence
Machine Learning
ARTICLE HISTORY
Received X XXX XXXX
Received in revised form
X XXX XXXX
Accepted X XXX XXXX
Available online X XXX XXXX
Artificial intelligence (AI) and machine learning have been revolutionizing
scientific research in recent years. As a result, chat-bot technology has made
significant strides in recent years, particularly with ChatGPT's emergence as a
prominent AI language model.[1] This in-depth analysis explores ChatGPT's
history, uses, major difficulties, and future plans. Before studying its numerous
uses in fields including customer service, healthcare, and education, first
explore its history, development, and underlying technology. It is also address
various mitigation measures while highlighting the significant difficulties that
ChatGPT confronts, such as ethical issues, data biases, and security
risks.Finally, by examining potential areas for more study and development,
emphasizing its integration with other technologies, enhanced human-AI
interaction, and bridging the digital divide, anticipate the future of ChatGPT.
For academics, developers, and other stakeholders interested in the rapidly
changing field of AI-driven conversational bots, this review provides insightful
information. This paper examines the numerous ways ChatGPT has transformed
scientific inquiry, from data processing and hypothesis formulation to
cooperation and public engagement.[2] The study also explores possible
difficulties and ethical issues related to the application of ChatGPT in research,
emphasizing the significance of finding a balance between AI-assisted
innovation and human knowledge. The study discusses a number of ethical
concerns with the current computing landscape and how ChatGPT can
challenge such ideas. The biases and restrictions of ChatGPT are also present in
this work. It is important to highlight that despite several issues and ethical
questions, ChatGPT has in a very short period of time gained amazing attention
from academics, research, and enterprises.

© 2022 The Authors. Published by Penteract Technology.

This is an open access article under the CC BY-NC 4.0 license
(https://creativecommons.org/licenses/by-nc/4.0/).
 1. INTRODUCTION
This Language models are becoming more complex
and diverse as a result of the quick development of artificial
intelligence (AI) and natural language processing (NLP). A
family of artificial intelligence models known as "generative
AI" may produce new data based on patterns and structures
discovered in previously collected data. These models are
capable of producing content in a variety of fields, including
writing, graphics, music, and more. Deep learning methods
and neural networks are used by generative AI models to
analyse, comprehend, and produce material that closely mimics
outputs produced by humans.[1] Among these, Open AI's
ChatGPT AI model has distinguished itself as a potent tool
with a wide variety of applications in diverse fields
Understanding ChatGPT's history and evolution is essential to
appreciate its contribution to furthering scientific research
[ The history, significant developments, and improvements Figure 1 : The stages of SDLC
achieved in the creation of ChatGPT are outlined in this part,
which also highlights the technological advancements that have During brainstorming meetings, ChatGPT opens up a world of
contributed to its success in the scientific community.[18] In possibilities for idea generation and requirements gathering. By
this regard, it is important to note that the ChatGPT is a conversing, we come up with new concepts and learn insightful
language model based on the Generative Pre-trained information.[2]
Transformer (GPT) architecture rather than a Generative
Adversarial Network (GAN) model.[2] GPT models are created Designing user-centric experiences has never been more
for natural language processing tasks like text production and enjoyable thanks to product prototyping! We work with
language interpretation, whereas GANs are often employed for ChatGPT to gain input on UX/UI decisions so that our designs
jobs like picture generation.The foundations of ChatGPT lie in are successful. We can create wireframe prototypes with
the study of NLP, an area of AI devoted to making it possible ChatGPT's capacity to decipher design descriptions, bringing
for robots to comprehend and produce human language.[19] A concepts to life.
very comprehensive and adaptable AI language model that ChatGPT provides important code company during the
could help with a variety of activities, such as text production, development process and aids with development and coding.
translation, and data analysis, was what motivated the ChatGPT enables our developers to provide high-quality
development of ChatGPT. The development of the Transformer solutions quickly, whether by providing code snippets or
architecture, which was first described in Reference [38], is the suggesting best practises.
basis of ChatGPT. It was created to address some of the Testing and QA: The latter just became more intelligent! We
drawbacks of earlier sequence-to-sequence natural language have conversations with ChatGPT about test plans and situations.
processing models, such as recurrent neural networks (RNNs) We make use of its knowledge to find suitable test cases and
and convolution neural networks (CNNs). Powerful language cutting-edge situations to raise the calibre of software.
models like OpenAI's GPT series, including GPT-2 and GPT-3, After-Launch Support and Maintenance: ChatGPT keeps
which served as forerunners to ChatGPT, were made possible improving even after its initial release. To solve problems and
by this ground-breaking architecture give our customers great service, we rely on its knowledge base.
. We expedite the maintenance procedure and maintain the highest
2. SOFTWARE DEVELOPMENT LIFE CYCLE level of software performance by talking with ChatGPT about
The software life cycle is represented pictorially improvements and problem fixes.[20]
and diagrammatically by a software life cycle model, Decision-making and Data Analysis:
also known as a process model. A life cycle model
Making decisions with data is simple! We use ChatGPT to
depicts every procedure needed to move a software
analyse project metrics and data and find insightful information.
product through each step of its life cycle. It also
We are able to optimise project outcomes by identifying patterns
captures the organizational framework in which these
and trends using its skills for exploratory data analysis.
techniques are to be used.[4] In other words, a life
cycle model depicts the different tasks carried out on a
piece of software from conception until retirement.
The essential development activities may be scheduled
according to phases in various life cycle models.
Therefore, no matter whether life cycle model is used,
all of the fundamental tasks are included, even if they
may be carried out in different sequences depending
on the life cycle model. Multiple activities can be
carried out at every stage of the life cycle.[13]
*Corresponding author:
E-mail address: Corresponding author <XXX@gmail.com>.
2785-8901/ © 2022 The Authors. Published by Penteract
Technology.
This is an open access article under the CC BY-NC 4.0 license
(https://creativecommons.org/licenses/by-nc/4.0/).
 Corresponding author et al./ Malaysian Journal of Science and Advanced Technology
the head margin in this template measures
proportionately more than is customary. This
measurement and others are deliberate, using
specifications that anticipate your paper as one part of
the entire proceedings, and not as an independent
document. Please do not revise any of the current
designations.[4]
Figure 2: Seven SDLC phases.
Requirements /Planning :Creating rapid prototypes of
complicated apps using ChatGPT may greatly ease
the requirements gathering step. Additionally, since
the analyst and client may agree on the prototype
before moving on to the construction phase, it can
reduce the likelihood of process
misunderstanding.[14]
Design: OpenAI's DALL-E deep learning model,
which produces digital pictures from descriptions in
natural language, can help with application design. In
the future, it may also be used to make sure that a
particular application's design complies with
regulatory standards like accessibility in addition to
offering user interface (UI) templates for typical use
cases.[16]
Build: ChatGPT is able to produce code in a variety
of languages. Small pieces of code might be written
as a support for developers, increasing their output
and the calibre of the software. It even has the
potential to make it possible for non-programmers to
write code.
Test: ChatGPT plays a significant part in the testing
process. It may be used to create different test cases
and to test the application simply by issuing natural
language inquiries. It may be used to do chaotic
testing to replicate worst-case scenarios to verify the
integrity of the application more quickly and
affordably, as well as to correct any vulnerabilities
that might be found using procedures like Dynamic
Code Analysis (DCA).
Maintenance: By assisting customers with
straightforward inquiries, ChatGPT may greatly
increase First Contact Resolution (FCR). By doing
this, it guarantees that issue resolution times are
drastically shortened and frees up support workers to
concentrate their attention selectively on issues that
are more complicated.
Users must be aware that security and privacy issues
2
with the present version of ChatGPT still need to be
adequately addressed, even if it has a significant role to
play in automating more cognitive processes in the
SDLC.[5]
When used properly, the SDLC offers development
teams a variety of benefits.clearly stated objectives The
danger of wasting time and money is reduced when
developers are crystal clear about the objectives they
must reach and the deliverable they must provide by a
certain deadline.The SDLC models use checks and
balances to guarantee that all programme is tested
before being put in larger source code.Stage-by-stage
progression Developers are unable to go on to the
following age until the previous one has been finished
and approved by a management.Member Modification
due to the well-organized project goals and
methodology papers included in SDLCs, team
members may leave and be replaced by new ones
rather smoothly.You Can Achieve Perfection
ChatGPT has a number of significant advancements
and enhancements over older generations, including.
(i) Improved context understanding: ChatGPT is more
successful at producing correct and pertinent text
because it can perceive and react to complicated and
nuanced inputs better .
(ii) Lessened biases: ChatGPT benefits from
continuing attempts to reduce biases in training data,
resulting in more objective and balanced outputs even
while biases are still present .
(iii) Capabilities for fine-tuning: ChatGPT may be
customised to meet the specific demands of researchers
in a variety of scientific fields by being fine-tuned for
certain activities and applications.
3. RESULT AND DISCUSSION
The higher bar set by ChatGPT for user
expectations must be taken into account by developers.
Your app's keyword search box needs to be upgraded
because it doesn't react with satisfying results and isn't
personalised. Customers and staff will anticipate
applications that answer inquiries and AI search
experiences with natural language queries as more
people become awed by ChatGPT's capabilities.
Modern software developers use generative AI in
finding coding samples or enhancing code quality are
two simple examples of how it might be beneficial.[17]
However, before integrating generative AI into their
application, product managers and their agile
development teams should evaluate and test their use
cases.ChatGPT, the cutting-edge AI chatbot that has
taken the globe by storm, has the ability to speed up
several phases of the Software Development Lifecycle
(SDLC), from requirements gathering through design
and testing, among other advantages.The field of low-
code/no-code and AI-assisted application development
has advanced significantly, but ChatGPT has the
potential to do much better. This effective AI
technology may be used to speed up numerous
processes at various stages of the SDLC, resulting in
shorter development cycles, increased developer
productivity, and quicker value delivery to businesses.
Benefits of ChatGPT.It offers several benefits that
make it a valuable tool in various application.Natural
 Corresponding author et al./ Malaysian Journal of Science and Advanced Technology
Language Understanding: ChatGPT is made to
comprehend natural language and provide text
answers that are human-like. It has the ability to
understand complicated questions, evaluate context,
and produce logical and pertinent responses. The
capacity to grasp natural language enables more
intuitive and engaged communication.[21]
ChatGPT has access to a broad variety of knowledge
and information since it has been trained on a
substantial amount of text data from the internet. It
can answer queries, clarify ideas, and provide
perceptions on a wide range of subjects.Real-time
support is made possible by ChatGPT's ability to
respond quickly. Users may get immediate feedback,
direction, and recommendations, which boosts
productivity and cuts down on support wait times.
Scalability and Availability: ChatGPT is a highly
scalable AI model since it can handle several
interactions at once. It can serve several users at once,
guaranteeing availability and responsiveness even
during busy times.
Language Flexibility: ChatGPT supports various
languages and can interpret and produce text in them,
making it available to a worldwide audience. Despite
linguistic difficulties, it makes support and
communication easier.
REFERENCES
[1] Hill-Yardin, Elisa L., Mark R. Hutchinson,
Robin Laycock, and Sarah J. Spencer. “A
Chat(GPT) about the Future of Scientific
Publishing.” Brain, Behavior, and Immunity
110 (May 2023): 152–54.
https://doi.org/10.1016/j.bbi.2023.02.022.
[2] Tuteja, Maneela, and Gaurav Dubey. “A
Research Study on Importance of Testing and
Quality Assurance in Software Development
Life Cycle (SDLC) Models” 2, no. 3 (2012).
[3] Amlani, Radhika D. “Advantages and
Limitations of Different SDLC Models.”
International Journal of Computer
Applications, 2012.
[4] Madapudi, Rudra Kumar. “Change Requests
Artifacts to Assess Impact on Structural
Design of SDLC Phases.” International
Journal of Computer Applications 54 (n.d.).
[5] “ChatGPT and Other Large Language Models
Are Double-Edged Swords | Radiology.”
Accessed June 24, 2023.
https://pubs.rsna.org/doi/abs/10.1148/radiol.2
30163?journalCode=radiology.
[6] Rahman, Md Mostafizer, and Yutaka
Watanobe. “ChatGPT for Education and
Research: Opportunities, Threats, and
Strategies.” Applied Sciences 13, no. 9
(January 2023): 5783.
https://doi.org/10.3390/app13095783.
[7] [2303.04671] Visual ChatGPT: Talking, Drawing
and Editing with Visual Foundation Models.”
Accessed June 24, 2023.
3
Support for Complex activities: ChatGPT can help
users with complex activities by offering instructions,
clarifications, and code hints. It can aid with code
debugging, syntactic support, and algorithmic
explanations, for instance, in software
development.Collaboration and Knowledge Exchange:
ChatGPT encourages communication and information
exchange among local communities. Users may
converse, exchange knowledge, and learn from one
another, fostering a cooperative and encouraging
atmosphere.
4. CONCLUSION
In conclusion, ChatGPT is an effective tool that can
be used to speed up and automate testing as well as
other software development operations. It might
completely change the way applications are created
while yet giving developers complete control over
their end result. Additionally, when it comes to
translating software requirements into actual code,
AI falls short of human efforts.[20] Certain
employment functions can and will be replaced by
ChatGPT, but the consequences won't be disastrous.
The secret is to comprehend the tool's potential and
how to use it, and to develop within our
responsibilities. Organizations are motivated by
profits and will seek to maximize them.
https://arxiv.org/abs/2303.04671.
[8] [2304.07590] Self-Collaboration Code Generation
via ChatGPT.” Accessed June 24, 2023.
https://arxiv.org/abs/2304.07590.
[9] [2306.07557] Ethical Aspects of ChatGPT in
Software Engineering Research.” Accessed June 24,
2023. https://arxiv.org/abs/2306.07557.
[10] Sallam, Malik. “ChatGPT Utility in Healthcare
Education, Research, and Practice: Systematic
Review on the Promising Perspectives and Valid
Concerns.” Healthcare 11, no. 6 (March 19, 2023):
887. https://doi.org/10.3390/healthcare11060887.
[11] Education in the Era of Generative Artificial
Intelligence (AI): Understanding the Potential
Benefits of ChatGPT in Promoting Teaching and
Learning by David Baidoo-Anu, Leticia Owusu
Ansah :: SSRN.” Accessed June 24, 2023.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id
=4337484.
[12] Ragunath, PK, S Velmourougan, P Davachelvan, S
Kayalvizhi, and R Ravimohan. “Evolving A New
Model (SDLC Model-2010) For Software
Development Life Cycle (SDLC),” 2010.
[13] Kumar, Naresh, A S Zadgaonkar, and Abhinav
Shukla. “Evolving a New Software Development
Life Cycle Model SDLC-2013 with Client
Satisfaction” 3, no. 1 (2013).
[14] Firat, Mehmet. “How Chat GPT Can Transform
Autodidactic Experiences and Open Education?”
Preprint. Open Science Framework, January 12,
2023. https://doi.org/10.31219/osf.io/9ge8m.
[15] Department of Computer Science & Engineering,
Noida Institute of Engineering & Technology
(NIET), Greater Noida, India., and Tanu Jindal.
“Importance of Testing in SDLC.” International
Journal of Engineering and Applied Computer
Science 01, no. 02 (December 31, 2016): 54–56.
https://doi.org/10.24032/ijeacs/0102/05.
[16] International Journal of Advanced Research in
 Corresponding author et al./ Malaysian Journal of Science and Advanced Technology 4

Computer Science and Software Engineering, n.d.

[17] Biswas, Som S. “Role of Chat GPT in Public
Health.” Annals of Biomedical Engineering 51, no.
5 (May 2023): 868–69.
https://doi.org/10.1007/s10439-023-03172-7.
[18] Biswas, Som. “Role of ChatGPT in Computer
Programming.: ChatGPT in Computer
Programming.” Mesopotamian Journal of
Computer Science 2023 (February 5, 2023): 8–16.
https://doi.org/10.58496/MJCSC/2023/002.
[19] “Study and Analysis of Chat GPT and Its Impact
on Different Fields of Study by Dinesh Kalla,
Nathan Smith :: SSRN.” Accessed June 24, 2023.
https://papers.ssrn.com/sol3/papers.cfm?abstract_i
d=4402499.
[20] Surameery, Nigar M. Shafiq, and Mohammed Y.
Shakor. “Use Chat GPT to Solve Programming
Bugs.” International Journal of Information
Technology and Computer Engineering, no. 31
(January 28, 2023): 17–22.
https://doi.org/10.55529/ijitc.31.17.22.
[21] “What If the Devil Is My Guardian Angel:
ChatGPT as a Case Study of Using Chatbots in
Education | SpringerLink.” Accessed June 24,
2023.
https://link.springer.com/article/10.1186/s40561-
023-00237-x.
[22] McGee, Robert W. “What Will the United States
Look Like in 2050? A Chatgpt Short Story.”
SSRN Electronic Journal, 2023.
https://doi.org/10.2139/ssrn.4413442.
[23] Iqbal, Syed Zaffar, and Muhammad Idrees. “Z-
SDLC Model: A New Model For Software
Development Life Cycle (SDLC)” 3, no. 2 (2017).
[24] Tuteja, Maneela, and Gaurav Dubey. “A Research
Study on Importance of Testing and Quality
Assurance in Software Development Life Cycle
(SDLC) Models” 2, no. 3 (2012).
 Corresponding author et al./ Malaysian Journal of Science and Advanced Technology 5

[1] G. Eason, B. Noble, and I. N. Sneddon, “On certain integrals of

Lipschitz-Hankel type involving products of Bessel functions,” Phil.
Trans. Roy. Soc. London, vol. A247, pp. 529–551, April 1955.
(references)
[2] J. Clerk Maxwell, A Treatise on Electricity and Magnetism, 3rd ed., vol.
2. Oxford: Clarendon, 1892, pp.68–73.
[3] I. S. Jacobs and C. P. Bean, “Fine particles, thin films and exchange
anisotropy,” in Magnetism, vol. III, G. T. Rado and H. Suhl, Eds. New
York: Academic, 1963, pp. 271–350.
[4] K. Elissa, “Title of paper if known,” unpublished.
[5] R. Nicole, “Title of paper with only first word capitalized,” J. Name
Stand. Abbrev., in press.
[6] Y. Yorozu, M. Hirano, K. Oka, and Y. Tagawa, “Electron spectroscopy
studies on magneto-optical media and plastic substrate interface,” IEEE
Transl. J. Magn. Japan, vol. 2, pp. 740–741, August 1987 [Digests 9th
Annual Conf. Magnetics Japan, p. 301, 1982].
[7] M. Young, The Technical Writer’s Handbook. Mill Valley, CA:
University Science, 1989.

Please refer to published articles in

https://mjsat.com.my/ for final published version of
the manuscript. IEEE Conference Proceedings
template has been used as reference for the creation
of this live template for MJSAT.