Harnessing the Potential

of
Generative Ai
Research &
for
Publication
Kapil Gupta
Professor, University of
Johannesburg
Organization of the Presentation

• Introduction to 4IR and Generative Artificial Intelligence

• Prompt Engineering

• Ai Ethics

• Navigating Research and Publication with Ai [Hands-on/Demo]

 Industrial Revolutions [Industry 1.0 to 4.0]

1st Industrial Revolution: Steam Power- The first

industrial revolution was driven by the steam engine and saw
Digital Technology Interventions the introduction of mechanical processes into manufacturing
plants
2nd Industrial Revolution: Electricity- The second
industrial revolution saw the introduction of mass production
and assembly lines powered by electricity
1784 1870 1969 today
3rd Industrial Revolution: Computing- The third
industrial revolution saw the very first application of
1.0 2.0 3.0 4.0 electronics and IT to aid production and automation

4rth Industrial Revolution: Digitalization- The

fourth industrial revolution has introduced the pervasive use of
digital technologies (cyber-physical systems and IoT etc.) to
manage, track, and monitor every aspect of the business
Mass processes.
Mechanization Computer Cyber-
Production
Water Power and Physical
Steam Power
Assembly Line
Automation Systems
Its not the Technology itself that
Electricity
brings about changes,
It is how the Technology is
applied
The phrase Fourth Industrial Revolution was first introduced by Klaus Schwab* in 2015.
*Founder and Executive Chairman of the World Economic Forum.

"Mastering the Fourth Industrial Revolution" was the theme of the

World Economic Forum Annual Meeting 2016 in Davos-Klosters, Switzerland. On October 10,
2016, the Forum announced the opening of its Centre for the Fourth Industrial Revolution in
San Francisco.

http://www.mum-writes.com/2018/06/rex-facing-the-4th-industrial-revolution-with-holistic-learning/

Mass Production and Customization to Mass Personalization

 Generative Artificial Intelligence (Ai)

John Mccarthy
American computer and cognitive scientist

Coined the term ‘Artificial Intelligence’ in 1956

Generative Ai is artificial intelligence capable of generating text,

images, videos, or other data using generative models, often in
response to prompts.
 Generative Ai is a type of artificial intelligence technology that
can produce various types of content, including text, imagery,
audio, and synthetic data.
Generative Ai
Also called Strong or Creative Ai
Traditional AI Generative Ai is able to create something new.
Also called Weak/Narrow/ or Restricted It focuses on creating new content like text,
AI images, audio and video, and other types of
focuses on performing preset tasks using contents.
predetermined algorithms and rules. These Smart, responsive and more powerful
artificial intelligence applications are It does not have any consciousness, emotions,
designed to excel in a single activity or a subjectivity
restricted set of tasks, such as playing chess, But can do sentiment analysis (opinion mining
diagnosing diseases, or translating i.e. identifying emotional tone behind a body
 Thelanguages.
main difference between traditional AI and generative AI lies in their of capabilities
text) and application.
Traditional
Smart,AI systems
but not are primarily responsive
necessarily used to analyze data and make predictions, while Generative AI goes a
step further by creating new data similar to its training data.
 Traditional AI excels at pattern recognition, while generative AI excels at pattern creation. Traditional AI can
analyze data and tell you what it sees, but generative AI can use that same data to create something entirely
new.
 Ai When user feed a Prompt then AI generator
processes the request and delivers the desired
Generators text, image, audio or other output. These
generators are a form of generative AI and use
large language models trained on millions of
data points to understand different prompts.
Work based on Large large language models trained on millions of data points (huge sets of
data) [hence the name ‘large’], reference this data to process a
Language Models
request, and produce the desired result.
(LLMs) A large language model (LLM) is a deep learning algorithm that can perform a variety of
natural language processing (NLP) tasks.
ChatGPT- 4
Elicit Typeset
SORA Quill Bot
Google Gemini
Research Rabbit Grammarly
DALL E 3 Pexels
https://openart.ai/home Connectedpapers
ZeroGPT Freepik
Consensus
Invideo
Paperdigest Designrr
 ChatGPT is a specific implementation of GPT that has been optimized for conversational AI. It
is designed to generate text that sounds like natural human speech, making it ideal for
chatbots, virtual assistants, and other conversational interfaces. ChatGPT has been pre-trained
on a massive amount of text data, including books, articles, and web pages, so it has a vast
knowledge base to draw from, when generating responses.

GPT (Generative­Pre-trained
Transformer)
Generative Pre-trained Transformer (GPT) is a type of neural network
architecture widely used in natural language processing tasks. It was first
introduced by OpenAI in 2018 as a way to generate high-quality, human-
like text. GPT is unsupervised learning, meaning it can learn from large
amounts of text data without direct human input.
ChatGPT is perhaps the most widely known generative AI tool at the moment. It is an AI
generator based on what is called a ‘large language model’, which combines deep-
learning algorithms with large data sets to transform clusters of information into
contextually and factually specific answers. This is based on human neural networks and
the way that humans process, sequence, and transform information into knowledge.
GPT- analyzes natural language queries, known as prompts, and predicts
the best possible response based on its understanding of language. To
do that, the GPT models rely on the knowledge they gain after they're
trained with hundreds of billions of parameters on massive language
datasets.
 Prompt Engineering
PROMPT is a mode of interaction between
a human and a large language model that
lets the model generate the intended output.
PROMPT ENGINEERING is a process of
creating input (usually text)
instructing the Generative AI to
generate the
Role, desiredInstructions
Task, Requirements, response
Preface for a Book
As an Editor of a Book, Write a Preface or propose a TOC
for a Book on Sustainable Quantum Computing, that
should be 700 words long, and Please include the
aspects related to SDGs

Project Proposal
As a PhD Student, please help me to draw the outlines of
a Project Proposal of 1000 words on ‘Machine Learning
for Commerce and Accounting’, that should not have
more than 4 sections, and introduction should be 300
words. Please also include the list of 10 most suitable
1. APE: Action, Purpose, Expectation
Prompt 4. CARE: Context, Action, Result, Example
•Action: Define the job or activity to be done.
•Context: Set the stage or context for the
•Purpose: Discuss the intention or goal. Engineerin discussion.
•Expectation: State the desired outcome.
Example g •Action: Describe what you want to be done.
•Result: Describe the desired outcome.
Framewor
A: Develop a Review Paper Outline on ‘3D Printing of •Example: Give an example to illustrate your point.
Graphene’
P: Our aim/purpose is to facilitate scholars
ks Example
C: We are working on a project related to
E: They should gain the knowledge and understand the
developing a small FFF based 3D printer
amount of previous research work done on this topic so far
2. ERA: Expectation, Role, Action A: We want to get the details of material
•Expectation: Describe the desired result. requirements, list of parts, power source capacity
•Role: Specify the role of ChatGPT. details, and manufacturing steps involved
•Action: Specify what actions need to be taken. for R: Our desired outcome is to achieve low operating
Example ChatGPT/D cost by running the 3D Printer using natural energy
E: It is expected to develop a solar panel made of functional E: A good example can be solar/wind power-based
material eepSeek devices
5. RACE: Role, Action, Context, Expectation
R: As an expert of 3D printing
•Role: Specify the role of ChatGPT.
A: To achieve this, Pls provide the most suitable and
•Action: Detail what action is needed.
efficient fabrication technique procedure
3. TAG: Task, Action, Goal •Context: Provide relevant details of the situation.
•Task: Define the specific task. •Expectation: Describe the expected outcome.
•Action: Specify the action required to address theGROK
task. 3 Example
•Goal: Clarify the desired outcome or goal. R: As a Sustainability Expert
(xAI) A: Compare all suitable techniques of 3D printing
Example
T: Fabricate an implant from functional material Qwen 2.5 C: To fabricate a functional material-based hybrid
A: Choose the most suitable 3D Printing technique(Alibaba)
based composite for high strength and corrosion free
on sustainability applications
G: The end goal is to cut down the implant cost and E: The expected outcome is to support circular
enhance properties compared to traditional implants economy principle and strengthen all three pillars of
 Ethical Implications
Possible misuse of Generative Ai –
Plagiarism and other academic misconduct

Possible concerns related to Generative Ai usage

Undermining the development of Critical Thinking, Writing Ability, and
Comprehension Skills

1) AI literacy; 2) AI ethics; and 3) AI skills

Proper Acknowledgment Should Be Given To The Ai

Resources Used
Do NOT share secrets with
Refrain from using Ai tools to bypass ChatGPT
PLAGIARISM DO NOT trust ChatGPT
blindly
Generative Ai for:

-Generating Ideas
Use generative AI to brainstorm ideas, get creative prompts, or test
assumptions

-Learning and Studying

Use generative AI to test yourself or a friend, deliver detailed summaries,
explain calculations, and provide activities for studying or research

-Feedback and Evaluation

Use generative AI to provide feedback on written work, evaluate assignment
answers, improve translations, and review your ideas
Limitations-
-Unverified and factually incorrect, nonsensical, or fabricated information
(increased factual accuracy GPT-4 https://openai.com/gpt-4)

-Errors in generated response against the prompt given

Successive prompts based on prompt engineering are required to be given to refine the
results

Benefits-
-Can reduce the admin tasks related to research & publications significantly
-Cost and Time Saving
Generative Ai for Research and
Publication
(how to navigate Research with Ai)
Let’s do it