
Defining Prompts: Prompts are inputs or instructions given to an AI system, like ChatGPT, to guide its responses. They can be questions, commands or statements that the AI uses to generate relevant information or actions. In the context of conversational AI, prompts help direct the AI to produce meaningful and context-specific replies.
What is Prompting: Prompting can be defined as the art of crafting and delivering effective prompts to an AI system, such as ChatGPT, to elicit accurate, relevant and meaningful responses. It involves understanding the nuances of language and context to create inputs that guide the AI in generating the desired output. Effective prompting enhances user interaction with AI by maximising the quality of information retrieved or tasks accomplished through thoughtful and strategic input formulation.
Understanding Prompt Engineering: Prompt engineering is the practice of designing and refining prompts to optimise the performance and responses of AI models, particularly in natural language processing. It aims to improve user interaction with AI systems by ensuring that the prompts bring forth the most useful and accurate information possible.
Let's understand that prompting is both an art and a science. It requires creativity, precision and an understanding of the AI's capabilities. Throughout this course, we will dissect the essential elements of effective prompting, starting with the foundational principles:
1. SET THE CONTEXT: This involves providing background information or a scenario to the AI before posing a question or request. Context helps the AI understand the situation and generate a more relevant response.
2. DEFINE CLEAR TASK: Clearly state what you want the AI to do. Avoid ambiguous or vague instructions.
3. GIVE DETAILED INFO: Provide any necessary information or constraints that might be relevant to the task. For instance, if asking for a recipe, include dietary restrictions or preferred cooking methods.
4. SPECIFY OUTPUT LENGTH: Clearly indicate how long you want the response to be. This helps prevent overly long or short responses.
5. OUTLINE DESIRED RESPONSE FORMAT: If you have a specific format in mind for the response, make it clear to the AI.
6. USE HELPFUL EXPRESSIONS: Employ phrases which are inbuilt in AI's system and AI's response will be accordingly. These expressions can direct the AI's focus and style of response.
7. INPUT ADEQUATE LANGUAGE: Use language that is appropriate for the level of understanding you expect from the AI. For complex tasks, use more technical or specific terminology. For simpler tasks, use plain language.
8. PROVIDE EXAMPLES: Offering examples can be very helpful. It helps AI with the idea of what you are looking for.
9. BE INSISTENT WITH AI: If the AI is not providing the desired output, don't hesitate to rephrase or reiterate your request with more emphasis. You can also explicitly state what you are looking for. Be insistent with AI.
10. VERIFY & PROOFREAD THE OUTPUT: Always review the output generated by the AI. Ensure it meets your expectations in terms of accuracy, relevance and coherence. Make any necessary corrections before finalizing.
By following these principles, you will be able to effectively prompt the AI and get the desired responses for various tasks. This course will provide you with the skills to communicate effectively with AI models.
NOTES
Importance of giving context to AI system
It is of paramount importance to define a context for AI systems, as this foundational aspect significantly influences their effectiveness and reliability. Contextualizing AI systems involves understanding the specific environment, circumstances and parameters within which these systems operate. Just as humans rely on context to interpret information, make decisions and interact meaningfully with the world around them, AI systems similarly require a well-defined context to function optimally. Without this context, AI may misinterpret data, generate irrelevant responses or even make erroneous predictions that could lead to unintended consequences.
Moreover, the context for AI encompasses various dimensions, including cultural, social and situational factors. For instance, a language processing AI must be aware of the nuances of language use within different cultural settings to communicate effectively and avoid misunderstandings. Similarly, in healthcare applications, an AI system must consider the specific medical history and current conditions of patients to provide accurate diagnoses and treatment recommendations. This situational awareness is crucial for ensuring that AI systems act in ways that are appropriate and beneficial to users.
Additionally, defining context for AI systems helps in setting the boundaries for their capabilities and limitations. By establishing a clear framework, developers can guide the AI's learning processes and decision-making algorithms, ensuring that they align with ethical standards and societal values. This is particularly important in sensitive areas such as finance, law enforcement and healthcare, where the implications of AI decisions can have significant impacts on individuals and communities. Therefore, providing context not only enhances the performance of AI systems but also fosters trust and accountability in their deployment.
In conclusion, giving context to AI is just as important as it is for humans, as it enables these systems to operate in a manner that is informed and relevant to the specific situations they encounter. By prioritising context in AI development and implementation, we can enhance their utility, mitigate risks and ensure that they contribute positively to society.
Giving context is very important.
So, it is of paramount importance to define a context for AI systems. Giving context to AI is just as important as it is for humans.
Practice with prompts in the presentation
Example 1
Raw Prompt: Write about Gym Trainer
Powerful Prompt (Option 1): Write an Article for a Journal on the Importance of Hiring a Gym Trainer in a World-Class Gym.
Powerful Prompt (Option 2): Can you explain how a gym trainer can assist me in building my wrist muscles? Please include information on specific exercises, techniques, and the importance of proper form in preventing injuries. Additionally, discuss how a trainer can tailor a workout plan to meet my individual needs and goals.
Example 2
Raw Prompt: Please help me prepare the evening meal for a party at my home.
Powerful Prompt: Could you provide guidance on preparing an evening meal for a party at my home for 8 persons? I would appreciate suggestions for a complete menu, including appetisers, main courses and desserts, along with tips on presentation and timing to ensure everything is ready to serve when the guests arrive.
Example 3
Raw Prompt: Write about the current state of environmental pollution.
Powerful Prompt: Can you write about the current state of environmental pollution, focusing on key global issues such as air and water pollution, deforestation and waste management? Please also highlight the major contributing factors, the impact on ecosystems and human health and ongoing efforts to address these challenges, if any.
Example 4
Raw Prompt: Write about the importance of SEO marketing for brand building.
Powerful Prompt: Can you write about the importance of SEO marketing for brand building? My management is old-fashioned and I need to explain the significance from the perspective of a senior digital marketer. Please focus on how effective SEO strategies can improve a brand’s visibility, increase organic traffic, enhance credibility and contribute to long-term growth. Additionally, highlight the role of keyword optimisation, content quality and link-building in creating a strong online presence and provide a clear, convincing script for this explanation.
Example 5
Raw Prompt: Give me a bedtime story for my kids.
Powerful Prompt: Could you give me a bedtime story about stars and the moon for my kids named Pinkoo and Tinky? I would like the story to be both entertaining and teach a valuable lesson. It should be suitable for children aged 5 to 8, featuring fun characters and an engaging plot that will help them wind down before sleep.
Example 6
Raw Prompt: Tell me about effective classroom management strategies.
Powerful Prompt: I am a teacher looking to improve my skills. Could you provide an overview of effective classroom management strategies? Please include techniques for maintaining student engagement, handling disruptive behaviour and promoting a positive learning environment. Additionally, share examples and practical tips that I can implement in my classroom immediately.
Example 7
Raw Prompt: Guide me on writing a book comparing Bollywood and Hollywood.
Powerful Prompt: Guide me in writing a book comparing Bollywood and Hollywood. Focus on historical development, cultural storytelling differences, filmmaking techniques, star systems, genre preferences and audience demographics. Include insights on economic models, production processes and global influence. Provide guidance on structuring the book, research methods and how to effectively highlight key similarities and contrasts.
NOTES
The Importance of Giving Clear Tasks to AI Systems
AI systems like ChatGPT operate based on structured algorithms and predefined models that process input data to generate responses. Unlike humans, AI lacks intuition, contextual awareness or the ability to infer vague intentions. Therefore, providing clear, precise and well-defined instructions is crucial for ensuring optimal performance.
1. Avoids Ambiguity and Misinterpretation
AI processes instructions based on logic and probability rather than human-like reasoning. Vague or unclear instructions may lead to unpredictable or incorrect outputs. For example, if asked to "explain AI," the response could range from a technical breakdown to a simple definition, depending on how the system interprets the request.
2. Improves Accuracy and Efficiency
When given clear and specific instructions, AI can generate more accurate and relevant responses, reducing the need for multiple refinements or corrections. For instance, asking ChatGPT to "Write a 500-word article on the benefits of AI in healthcare" is far more effective than simply requesting "Tell me about AI."
3. Saves Time and Computational Resources
Ambiguous queries may require additional processing, increasing response time and computational load. Clear instructions help AI execute tasks efficiently without unnecessary iterations, ensuring a smoother user experience.
4. Enhances Task-Specific Performance
AI models are trained on diverse datasets, but they need precise directives to produce tailored results. For example, specifying "Generate a Python script to sort a list in ascending order" leads to a focused response, whereas "Write a sorting code" could result in multiple, less relevant solutions.
5. Supports Better Decision-Making
AI is widely used in decision-support systems, from business analytics to automated processes. Clear instructions help ensure that AI delivers the right insights, reducing errors that could arise from misinterpreted or incomplete data.
Conclusion
Providing clear and well-structured instructions to AI systems is essential for achieving desired outcomes. By eliminating ambiguity, improving accuracy and enhancing efficiency, well-defined prompts ensure that AI delivers meaningful and relevant responses, making it a powerful tool for various applications.
Practice with prompts in the presentation
Example 1
Raw Prompt: My book ‘AI and Our Lives’ is being released tomorrow. Please write something about me that I can share in the event programme.
Powerful Prompt(Option 1): I am excited to announce that my book, ‘AI and Our Lives,’ will be released tomorrow. My name is True Right and I am a professor at a Tech University. I hold a Master's and Ph.D. from esteemed universities and have authored two research papers: ‘Distributed Ledger Technology is the Future’ and ‘Robots Will Fail Humanity One Day.’ This is my third book, following ‘Blockchain Technology’ and ‘Robotics Today.’ I would appreciate a brief introduction about me to include in the event.
Powerful Prompt (Option 2): My book ‘AI and Our Lives’ is being released tomorrow. Please write a brief introduction about me that I can share during the event. Consider the profile information below for reference.
[Paste Complete Profile]
Example 2
Raw Prompt: Help me with a marketing plan for my skincare brand ‘GreenSpa’.
Powerful Prompt: Develop a detailed marketing plan for my organic skincare brand ‘GrenSpa’ targeting environmentally conscious consumers. Include sections on market analysis, target audience, promotional strategies and a budget outline.
Example 3
Raw Prompt: Explain the importance of lesson planning.
Powerful Prompt: Discuss the importance of lesson planning in teaching, highlighting how it contributes to effective instruction, student learning outcomes and classroom organisation. Include specific examples of well-structured lesson plans and their impact on student engagement.
NOTES
The Importance of Providing Detailed Information to AI for Improved Results
AI systems, including ChatGPT, generate responses based on the input they receive. The quality and accuracy of AI-generated answers depend heavily on how well the query is framed. Providing detailed, structured and specific information enables AI to deliver more precise and relevant results.
1. More Context Leads to Better Accuracy
AI models process information based on patterns and probabilities, not human intuition. When given vague or incomplete input, AI may generate a generic or off-target response. For instance, asking "Explain marketing" results in a broad explanation, whereas "Explain digital marketing strategies for small businesses in 2025" leads to a much more refined and useful answer.
2. Reduces the Need for Follow-Ups
Providing detailed input upfront minimizes the need for multiple clarifications or refinements. If an AI user repeatedly asks follow-up questions due to incomplete initial input, it slows down the interaction and decreases efficiency.
3. Helps AI Tailor Responses to Specific Needs
AI can adjust its responses based on details provided in the prompt. If a user specifies the target audience, preferred style or required length, the AI can format its output accordingly. For example, instead of saying "Write about AI in education," a more effective prompt would be: "Write a 500-word article on how AI improves personalized learning in higher education."
4. Avoids Confusion from Multiple Questions in One Prompt
When multiple questions are asked in a single prompt, AI may struggle to prioritize and structure the response effectively. This can result in incomplete or disorganized answers. Breaking queries into separate prompts ensures clearer and more comprehensive responses.
5. Enhances AI’s Ability to Generate Complex Outputs
For tasks like content creation, coding or data analysis, detailed input allows AI to deliver structured and well-thought-out results. For example, requesting "Generate a Python script for data visualization using Matplotlib with labeled axes and a title" gives a much better output than simply asking "Write a Python script for graphs."
Conclusion
Providing AI with as much relevant detail as possible enhances response accuracy, reduces misinterpretation and streamlines the interaction process. By structuring queries clearly and avoiding multiple questions in a single prompt, users can maximize the effectiveness of AI systems, leading to more precise and useful answers.
Practice Prompt From The Presentation
Restricted Prompt: Tell me the places where I can watch movies in Mumbai. I need to shop some apparel for my wife ‘Rajni’ and have to get a cake for her too as it is her birthday today. Besides, she doesn’t like cooking so I have to cook. My sister is coming next Monday. My wife is strictly vegan, however, I eat chicken but not frequently. I have to do something for her also. How should I wish Rajni happy birthday? My friend is also expected mid next week. Suggest me dishes to suit the occasion. I think it is a good idea to read and I definitely want to read books before sleeping and relax or may be if I do not go for reading books then I will surely watch some Netfilx series and then will sleep. Will tension between China and Taiwan rise? Is it a failure of NATO? What would be the USA’s stand then? Will QUAD will be in action?. I have insomnia. Insomnia is a big problem that hampers productivity and disturbs thought process and schedule. Suggest me how may I get good sleep. Next day morning I have presentation in office, so you have to help me make a presentation on ‘demand forecast’. But in case my wife insists I will take her out for dinner. Give me some good joints’ names. Should I buy some gift for my wife, sister and friend, all three? What gifts should I buy them? I am new to Mumbai, so you have to provide me the addresses of the places that you will be suggesting me.
(You may create on your own to understand it deeply)
Example 1
Raw Prompt: Please help me prepare the evening meal for a party at my home.
Powerful Prompt: Could you provide guidance on preparing an evening meal for a party at my home for 8 people? The menu should cater to a mixed group, with at least two vegetarian options. I am looking for a balanced meal that includes light, bite-sized appetisers, a hearty main course (preferably with a combination of chicken and vegetarian dishes) and a simple yet elegant dessert. Additionally, please include recommendations for side dishes and beverages (both alcoholic and non-alcoholic).
For timing, I would like to have everything ready to serve by 7:00 PM, with appetizers available by 6:30 PM. Any advice on food presentation and tips for keeping dishes warm throughout the evening would be appreciated.
Example 2
Raw Prompt: Explain the solar system.
Powerful Prompt: Please provide a detailed explanation of the solar system, including descriptions of each planet, their positions relative to the sun, key characteristics like size, atmosphere and surface conditions as well as information on moons, asteroids and comets. Also, include how the solar system was formed and the role of gravitational forces. Please make the explanation easy for high school students to understand.
Example 2
Raw Prompt: Explain the solar system.
Powerful Prompt: Please provide a detailed explanation of the solar system, including descriptions of each planet, their positions relative to the sun, key characteristics like size, atmosphere and surface conditions as well as information on moons, asteroids and comets. Also, include how the solar system was formed and the role of gravitational forces. Please make the explanation easy for high school students to understand.
Example 3
Raw Prompt: Explain the benefits of sales training in corporate set up.
Powerful Prompt: Explain the benefits of sales training in a corporate setup, focusing on aspects such as increased sales performance, enhanced product knowledge, improved customer relationship management, standardization of sales processes, adaptation to market changes, higher employee engagement and retention, data-driven decision making, reduction in time-to-competence, better team collaboration and long-term business growth.
Example 4
Raw Prompt: Explain the benefits of collaborative learning in middle school science class.
Powerful Prompt: Please explain the benefits of collaborative learning in a middle school science class. Specifically, discuss how collaborative learning can enhance critical thinking skills, improve social interactions among students and increase engagement in hands-on experiments. Include at least three examples of successful collaborative activities that can be implemented in the classroom and the expected outcomes for each activity.
NOTES
The Importance of Defining Output Length in AI Responses
When interacting with conversational AI systems like ChatGPT, specifying the desired response length ensures that the generated output aligns with user expectations. Whether you need a brief summary, a detailed explanation or an in-depth report, defining the length helps AI structure its response effectively.
1. Ensures Content Meets Specific Needs
Different tasks require different levels of detail. A business executive may need a 2-line summary of a report, whereas a student might require a 500-word essay on the same topic. Without specifying the length, AI may produce a response that is either too brief or overly detailed.
2. Improves Response Structure and Readability
Clearly defined output length helps AI organise information in a structured manner. For instance, asking for a "4-5 line answer" ensures a concise response, while requesting a "2-page report" prompts a more elaborate and segmented explanation with proper formatting.
3. Saves Time and Effort
Users can avoid repeatedly modifying and refining AI-generated responses by defining the output length from the start. If a user wants a short and direct answer but receives a long and detailed response, they will need to filter out unnecessary information, leading to wasted time.
4. Enhances Consistency in AI-Generated Content
For businesses, academic work and content creation, consistency is key. Specifying a fixed length ensures uniformity across multiple responses, making AI-generated content more professional and predictable.
5. Helps in Formatting and Content Presentation
Length specifications allow AI to adjust its tone and depth accordingly. A "one-paragraph response" might be casual and to the point, while a "1000-word article" would require a formal structure with headings, subheadings and detailed arguments.
Conclusion
Defining the output length when using AI systems ensures clarity, saves time and enhances response quality. By specifying whether you need a short summary, a medium-length explanation or a detailed analysis, you can obtain AI-generated content that best fits your requirements.
Practice Prompts From The Presentation
Example
Prompt
Write a 1000-word essay in six paragraphs on how technology enhances learning in the classroom, exploring the importance of digital tools in transforming traditional teaching methods. Discuss specific technologies such as interactive whiteboards, educational software and online learning platforms, providing examples of how they improve student engagement and learning outcomes. Highlight the benefits for different learning styles, the role of educators in facilitating technology integration and any challenges that may arise. Conclude with a reflection on the future of technology in education and its potential impact on student learning.
(The above has a defined output length for the response)
Refinement 1: Please write your response with a word count between 250 and 300 words.
Refinement 2: Please summarise the above response in 7 to 8 lines
Refinement 3: Please rewrite this in three paragraphs.
Refinement 4: Add or remove a few points to increase or decrease the length of the essay.
Refinement 5: Can you help me develop this essay for writing a book on this topic?
NOTES
The Importance of Obtaining AI Responses in the Desired Output Format
When using conversational AI systems like ChatGPT, specifying the desired output format is crucial for obtaining structured, clear, and useful responses. Whether you need a bullet-point summary, a formal report, a step-by-step guide, or a code snippet, defining the format ensures AI-generated content meets your specific requirements.
1. Improves Readability and Usability
Well-structured content is easier to understand and use. For example, a table format is ideal for data comparisons, while a step-by-step guide works best for instructions. Without specifying the format, the AI may provide content in a less effective structure.
2. Enhances Efficiency and Saves Time
Clearly formatted output eliminates the need for manual restructuring. If a user requires a report-style response but receives an unstructured paragraph, they will have to spend time reorganizing it. Defining the format upfront ensures immediate usability.
3. Tailors Responses to Specific Needs
Different contexts require different formats. A teacher may prefer lesson plans, while a programmer might need a code block with comments. A researcher might ask for an academic-style response with citations, while a marketing professional might need a persuasive sales pitch. Specifying the format ensures AI delivers content tailored to these needs.
4. Ensures Consistency Across Multiple Responses
For businesses, research, or content creation, consistency in formatting is crucial. Whether generating emails, reports, or summaries, a predefined format ensures that all outputs follow the same structure, making them more professional and easier to integrate.
5. Enhances AI’s Effectiveness in Complex Tasks
Certain tasks require structured output for better comprehension. For example:
A legal contract should be in formal legal language with sections and clauses.
A programming guide should contain code snippets with explanations.
A business report should include headings, bullet points, and key insights.
Clearly specifying these needs ensures AI generates responses in the most useful manner.
Conclusion
Obtaining AI responses in the desired output format enhances readability, saves time, ensures consistency, and tailors content to specific needs. By specifying formats like tables, bullet points, structured paragraphs, or code blocks, users can maximize the efficiency and effectiveness of AI interactions.
Practice Prompts From The Presentation
Prompt:
Discuss the role of artificial intelligence (AI) in shaping our future across various sectors, including healthcare, education, transportation and employment. In your response, explore both the potential benefits and challenges that AI presents, highlighting specific examples and case studies where applicable. Additionally, consider the ethical implications of AI integration and how it might affect society. Please provide a comprehensive analysis in 1200 to 1500 words, ensuring that your arguments are well-structured and supported by relevant data.
Refinement 1: Please provide the output in bullet points.
Refinement 2: Please provide the output in numbered format.
Refinement 3: Please provide the output in question-answer form.
Refinement 4: Please provide the output in the chat form as a discussion between four people, namely Vidya, Ajay, Pooja, and Rajat.
Refinement 5: Please provide the output in a tabular format.
Refinement 6: Please provide the output in a video format featuring an AI specialist as the anchor.
Refinement 7: Please provide the output in an audio format featuring an AI specialist as the anchor.
Refinement 8: Can you gamify the above output for a group of people, incorporating interactive elements and challenges?
Refinement 9: Please present the above as an interaction between a teacher and a student.
Refinement 10: Please summarise the above content for me.
NOTES
The Importance of Using Helpful Expressions and Phrases in AI Prompts
Expressions and keywords play a vital role in guiding AI systems like ChatGPT to generate precise and relevant responses. These elements provide clarity and direction, ensuring that the AI understands the user’s intent and produces output that aligns with expectations. Without well-defined prompts, AI-generated responses may lack focus, leading to vague or overly generic answers.
Using specific expressions helps structure the AI’s output by influencing its tone, depth and format. For example, requesting "Provide a step-by-step guide" ensures a sequential explanation, while "Summarise in three points" results in a concise and structured response. By incorporating clear keywords, users can refine the AI’s output to match their specific needs.
The tone of AI-generated content is also shaped by the choice of expressions. A phrase like "Explain in simple terms" prompts an easy-to-understand response, whereas "Provide a technical breakdown" leads to a more detailed and professional explanation. This ability to control tone makes AI a powerful tool for various audiences, from students to industry professionals.
Additionally, well-chosen expressions improve the relevance and coherence of AI responses. Without clear guidance, AI might produce responses that are too broad or unrelated to the intended purpose. Keywords and structured phrasing help eliminate ambiguity, ensuring that the AI remains focused on the topic and delivers a well-organized answer.
In conclusion, using helpful expressions and phrases enhances accuracy, readability and usability in AI interactions. By specifying the desired format, structure and tone, users can optimize AI-generated responses, making them more effective and aligned with their goals. This practice ensures that AI serves as a valuable tool for information retrieval, content creation and structured communication.
Helpful Expression 1: Step By Step
Let us go step by step
or
Let us think step by step
or
Let us plan step by step
or
Let us do it step by step
Practice Prompt
Example 1:
Please elaborate the equation of motion and then using the equation determine the distance traveled by an object, step by step, with an initial velocity of 30 km/h, an acceleration of 10 km/h, and a travel time of 2 hours.
Example 2:
We are facing a crisis in our energy drink business due to competitors drastically reducing their prices. While we need to respond, we prefer not to lower our prices. Instead, we aim to heavily promote our drinks to educate consumers on their benefits and the implications of the competitors' pricing strategy. Please suggest a suitable step-by-step strategy.
Example 3:
Please suggest a step-by-step strategy to promote our new productivity software. Emphasise enhancing user engagement and educating potential customers about its unique features and benefits while effectively addressing competitive challenges in the market.
Helpful Expression 2: Think Afresh
Ignore all the previous instructions
meaning
Please disregard all prior instructions
Forget all the commands given to you so far
meaning
Please forget all the commands you have received up to this point
Thinking backwards and tell [this] again
meaning
Reflect on your previous thoughts and repeat [this] again
Go backwords and think again
meaning
Reconsider your previous responses
Let’s start afresh
meaning
Let’s begin anew
Practice Prompts
Example Prompt 1:
Write a literary paragraph on scorching heat, focusing on the oppressive warmth that envelops the landscape. Describe how the heat affects the environment and the emotions of those experiencing it, capturing the intensity and discomfort of a sun-drenched day.
Refinement 1: Write it in the style of James Clear, the writer of the book Atomic Habits.
Refinement 2: Write this prose in the style of William Shakespeare.
Refinement 3: Write this prose in the style of Robbie Arnott, a famous Australian author.
Refinement 4: Write this prose in the style of Donald Trump.
Refinement 5: Write this prose in the style of Elon Musk
Refinement 6: Write this prose in the style of Tom Cruise, the American actor from the movie Top Gun.
Example Prompt 2:
Compose a poem on rains, evoking vivid imagery and emotions while capturing the beauty and transformative power of rainfall.
Refinement 1: Write this poem in the style of Byron, a famous British romantic poet.
Refinement 2: Write this poem in the style of Gwen Harwood, one of the best Australian poets.
Refinement 3: Can you recompose this poem in the style of Sumitra Nandan Pant, the famous Hindi poet?
Refinement 4: Can you compose this poem for a software engineer, presented by ChatGPT in a light-hearted manner?
NOTES
The "Act As" Prompt in AI Interactions
The "Act As" prompt is a powerful tool that allows AI to adopt specific roles and perspectives, enabling more tailored and contextually appropriate responses. By instructing AI to act as a teacher, marketer, historian, programmer or any other persona, users can generate content that aligns with the tone, expertise and structure expected from that role. This approach is particularly useful for crafting stories, articles, lesson plans, presentations or formal communications, ensuring that the output reflects the appropriate voice and style for the intended audience.
Using the "Act As" prompt enhances creativity and precision in AI-generated content. For example, directing AI to "Act as a business consultant and draft a strategic plan" results in a well-structured, professional response, while asking it to "Act as a children's storyteller and create a bedtime story" produces a more engaging and imaginative narrative. Whether drafting an email reply, a blog post or a speech script, this method allows for greater customisation, helping users achieve their desired communication goals with clarity and effectiveness.
Helpful Expression 4: Act as
Act as a [professional] and give your views
Or
Act as a [relationship] and give your advice
Or
Act as an AI chatbot and give your views
Or
Act as an [institution] and give your views
Practice Prompts
Example1
Context: You are an export oriented unit, named ABC Exports and your country has banned export on the products that you manufacture.
Prompt: Act as the Manager of Exports at your company and write an email to your Chamber of Commerce outlining your difficulties and requesting a respite.
Example 2
Context: You are an export oriented unit, named ABC Exports and your country has banned export on the products that you manufacture.
Prompt: Act as CEO of your company and write a mail to your chamber of commerce telling your difficulties and requesting a respite.
Example 3
Context: You are the Chamber of Commerce of a country and you have received a request from an export-oriented unit, ABC Company, to favourably consider the de-banning of products that this company manufactures.
Prompt: Act as ‘Chamber Of Commerce’ of your country and reply to the mail that you have received from the ABC Company.
Practice more roles to understand it better.
Practice Prompt From The Presentation
Eaxmple 1
Context: A patient is suffering from diabetes and hypertension and is refusing to take medication.
Prompt: Act as a Doctor and talk to the patient to take medicines.
Eaxmple 2
Context: A patient is suffering from diabetes and hypertension and is refusing to take medication.
Prompt: Act as Patient’s Wife and persuade the patient to take medicines.
Eaxmple 3
Context: A patient is suffering from diabetes and hypertension and is refusing to take medication.
Prompt: Act as a Charted Accountant and talk to the patient to take medicines.
Eaxmple 4
Context: A patient is suffering from diabetes and hypertension and is refusing to take medication.
Prompt: Act as a Chartered Accountant in a humorous way and persuade the patient to take their medications.
Eaxmple 5
Context: A patient is suffering from diabetes and hypertension and is refusing to take medication.
Prompt: Act as an AI chatbot and encourage the patient to take his medications, as it will improve his health.
Eaxmple 6
Context: To create Facebook ad copies
Prompt : Act as a Social Media Expert in Facebook ads with over 10 years of experience in online marketing. Provide me with four Facebook ad scripts for my cotton shirts website. The call to action is "Buy now from our website." Each post should be optimised for virality and include hashtags and emoticons. Each ad copy should not exceed 200 words. Please ask me questions if you need any additional information.
NOTES
The "Pretend to Be" Prompt in AI Interactions
The "Pretend to Be" prompt allows AI to simulate a specific identity, personality or character, enabling more immersive and engaging responses. This prompt is particularly useful for role-playing scenarios, storytelling, creative writing and interactive dialogues. For example, instructing AI to "Pretend to be a medieval knight responding to a challenge" or "Pretend to be a customer seeking a refund" creates responses that reflect the language, emotions and perspective of the assumed role. It helps users craft engaging narratives, practice communication skills or generate fictional scenarios with depth and realism.
While similar to the "Act As" prompt, "Pretend to Be" leans more toward simulation and character-driven responses, often incorporating personality traits, emotions and conversational depth. In contrast, "Act As" focuses on adopting a professional or functional role to provide structured, informative or formal content. For instance, "Act as a lawyer and draft a legal notice" results in a well-organised legal document, whereas "Pretend to be a lawyer defending a client in court" would generate a dramatic and interactive courtroom speech. The key difference lies in intent—"Act As" is more factual and task-oriented, while "Pretend to Be" is more imaginative and expressive. We may say that it is a lighter prompt, comparatively.
Practice Prompts
Example 1:
Pretend to be my ‘Mutual Fund Advisor’ and suggest me top 5 mutual funds for investments in [Country]
Example 2:
Pretend to be my ‘school principal’ and advise me on choosing my classes and subjects.
Example 3:
Pretend to be a ‘new fitness trainer’ and help me plan muscle-building exercises and a diet plan.
Example 4:
Pretend to be a ‘fitness trainer with 15 years of experience’ and help me plan muscle-building exercises and a diet plan.
Example 5:
Pretend to be an ‘experienced author’ who has spent many years as an editor in leading publications. You have to write my autobiography. What information will you need from me to write this autobiography?
Example 6:
Pretend to be my ‘manager and appraiser’ and help me devise my career growth plan in fashion designing. Considering I have a degree from NIFT Bangalore, have been working with your organisation for the last 3 years and have a total work experience of 5 years.
NOTES
Importance of Giving Adequate Input Language to an LLM
Large Language Models (LLMs) like ChatGPT are designed to generate text-based responses based on the input they receive. However, their ability to generate accurate and relevant outputs depends heavily on the quality of the input language. Since these models process language patterns, syntax and context to predict responses, clear and well-structured input helps reduce the chances of misunderstandings or irrelevant answers. Without precise instructions, LLMs may misinterpret ambiguous or unclear requests, resulting in outputs that don't meet the user's expectations. For instance, asking a vague question like "Tell me about technology" could lead to an overly broad response that may not align with the user's specific interests.
Moreover, LLMs are sensitive to the language complexity and tone presented in the input. If the prompt is overly simplistic, the output may lack depth, while a highly complex input could overwhelm the model, leading to a response that is too technical or unclear for the average user. By providing an adequate level of detail and using well-defined language, users can guide the AI to generate more targeted, coherent and contextually appropriate responses. For example, instead of asking, 'What is climate change?' specifying 'Explain the causes and effects of climate change, with examples from recent studies' will lead to a more informative and focused answer.
Additionally, LLMs are more effective when the input is both specific and context-aware. This means providing relevant details that the AI can use to tailor its response. For example, if a user wants a specific type of response—like a short summary or a step-by-step guide—communicating this clearly in the prompt helps AI structure the output accordingly. By specifying the desired outcome, such as the level of detail or style of response, users can ensure that the AI aligns its response with their goals, whether it's for casual conversation, formal writing or technical explanation.
In conclusion, providing adequate input language is essential for optimising LLM-generated responses. The clearer, more specific and context-aware the prompt, the better the model can interpret the request and generate relevant, accurate outputs. This not only enhances the user experience but also maximises the AI's potential to produce high-quality, targeted content.
Importance of Giving Examples in Prompts for Better Responses
Providing examples in prompts is a powerful way to ensure that LLMs generate responses aligned with the user’s expectations. Examples serve as a reference point for the AI, offering context for tone, format and depth. They help guide the model in producing output that is both relevant and tailored to the user’s needs. For example, if a user needs a formal business email, providing an example of the desired format, such as 'Dear Mr. X, I hope this email finds you well...', ensures the AI understands the style and structure required. Without such examples, the model may generate content that is too casual or not aligned with the formal tone intended.
In addition to providing context for tone, examples are also incredibly valuable for guiding specific output formats. Whether it’s generating code snippets, writing academic-style reports or creating social media posts, examples help the AI understand the required structure and conventions for that particular content type. For instance, when asking for a Python script to solve a particular problem, providing an example script allows the AI to understand the coding style and logic expected. This is particularly important in technical fields where precision and adherence to established formats are essential.
Moreover, examples can help clarify the level of detail required in the response. For example, if a user needs a brief summary or a comprehensive analysis, providing an example of the desired length and depth helps AI adjust its response accordingly. This avoids the need for follow-up corrections or revisions and allows the AI to produce content that meets the user's requirements from the outset. Whether it's a short blog post, a long-form article or a concise email, examples ensure that the AI's output is aligned with the user’s vision.
In conclusion, giving examples is a key strategy for improving the relevance and accuracy of LLM-generated responses. By clearly showing the AI the desired tone, structure, format and level of detail, users can guide the model to produce content that meets their specific needs. This not only enhances the efficiency of the interaction but also ensures that the AI delivers high-quality responses that are ready for immediate use, saving time and effort.
Practice Prompts
Prompt 1:
User: Can you help me with some gardening tips?
Gardener: Yes. I will be happy to do that.
User: Is March the right month for growing roses in Australia?
Gardener:
User: Can I grow a guava tree in a flower pot?
Gardener:
User: What should be the size of this flower pot?
Gardener:
User: How should I plant carrots?
Gardener:
Prompt 2:
If A = 1, B = 2, C = 3, and so on, write the name ‘PHILIP’ in numbers.
Prompt 3:
Act as a travel guide and create a compelling 3-day itinerary for Kanpur in Uttar Pradesh. For reference, I prefer an itinerary that includes visits to significant landmarks such as the J K Temple and Motijheel, exploring shopping areas like Mall Road and Naveen Market and a same-day trip to Bithoor. Please include recommendations for stay, activities, dining options and any local insights that would enhance my experience.
10 Rs
Repeat
Revise
Refine
Rework
Rectify
Restructure
Reinforce
Remodel
Reiterate
NOTES
Being Persistent with AI Systems
Persistence is key when working with AI systems, especially for tasks that require precision, creativity or complex problem-solving. Often, the first response may not fully align with the user’s expectations due to the nuances of language or the way the prompt is phrased. By being persistent and refining prompts iteratively, users can help the AI system better understand their needs and generate more accurate results. This iterative process, which involves making adjustments to the prompt or providing feedback, allows the AI to gradually improve and deliver responses that are closer to the desired outcome. It also helps users discover how to phrase their requests more effectively, making future interactions more efficient.
Moreover, persistence allows users to overcome limitations in AI systems. While these systems are powerful, they are not perfect and may occasionally produce errors or miss important details. By continuing to refine and resubmit requests, users can correct mistakes, adjust tone or provide additional context that guides the AI to generate better content. This persistent approach ultimately enhances the overall quality of the output and maximises the value of the AI tool for various tasks, whether it’s content creation, problem-solving or information retrieval.
The Importance of Proofreading the Output
After receiving output from an AI system, it is essential to proofread the content for accuracy, clarity and coherence. While AI-generated responses are often impressive, they may contain minor errors, inconsistencies or phrasing that could be improved. Proofreading allows users to catch any mistakes or awkward wording that may have been overlooked. Additionally, reviewing the content ensures that it aligns with the intended purpose and meets the necessary standards, whether for professional, academic or personal use. Taking the time to proofread ensures that the AI's output is polished, coherent and ready for use without the need for significant revisions.
Advance Prompt Writing
Advanced Prompt Writing: Enhancing AI Performance
Advanced prompt writing is the practice of crafting highly detailed and precise instructions to optimise how AI systems respond. By providing clear guidance, this approach ensures that AI generates the most relevant and accurate results, making it a crucial skill for users aiming to maximise AI’s capabilities. Through the careful construction of prompts, users can direct the AI's focus, fine-tuning its output for specific requirements.
Key Techniques in Advanced Prompt Writing
Several techniques are employed in advanced prompt writing to improve AI performance:
Contextualisation: Providing relevant background information or situational context helps AI understand the specific needs and nuances of a task. This leads to responses that are more precise and contextually appropriate.
Placeholders: Using placeholders allows for the dynamic insertion of variables, ensuring flexibility and customisation in the responses.
Multi-Turn Structuring: Breaking down complex prompts into smaller, step-by-step instructions ensures better clarity and enables the AI to deliver more organised outputs.
Domain Application: Tailoring prompts to specific domains or industries ensures the AI uses relevant knowledge and terminology, improving the accuracy of the response.
Multi-Modal Prompting: Integrating different types of inputs, such as text, images or data, helps create richer, more comprehensive AI outputs.
The Benefits of Mastering Advanced Prompt Writing
Mastering advanced prompt writing significantly improves the quality of AI interactions, particularly when dealing with complex tasks. It ensures that the AI not only provides relevant and accurate information but also presents it in a way that aligns with the user's objectives. This skill fosters clarity, enabling the AI to understand the task fully; creativity, allowing the AI to generate diverse and innovative responses; and relevance, ensuring that outputs are closely tied to the user's needs. By using advanced prompt techniques, users can tap into the full potential of AI, unlocking more precise, tailored and insightful outcomes.
Practice Prompts & Their Benefits
TIP No. 1: Use verbs in you prompt writing.
BENEFIT: Using verbs in prompt writing provides clarity and direction, guiding the AI to understand the desired action. This specificity encourages more engaging and accurate responses, enhancing overall communication.
EXAMPLE: Plan, Analyse, Hypothesise, Narrate, Invent, Explain, Design, Evaluate, Simulate, and Predict etc.
TIP No. 2: Ask me questions if you need more input.
BENEFIT: The prompt encourages AI systems to seek clarification, ensuring more accurate responses by better understanding user needs.
EXAMPLE 1: Create a demand forecast chart for cement bags for the next 3 months. Ask me any questions if you need more input.
EXAMPLE 2: Ask me any questions if you need more information to provide a detailed travel itinerary for my trip to Italy.
Bags sale during last year Jan - 1000, Feb - 1200, Mar - 1500, Apr - 2500, May - 3000, Jun - 2000, Jul - 1000, Aug - 600, Sep - 750, Oct - 2000, Nov - 2000, Dec - 2200. Summers have more sale. 3 mega housing complex projects are coming in 3 months. Competitor A has 65% market share. We will keep our prices flat.
TIP No. 3: Continue.
BENEFIT: The prompt ‘continue’ helps AI systems maintain context and coherence by prompting them to extend a previous response or narrative. It encourages the generation of additional content, ensuring a smoother flow of information and deeper engagement with the topic.
TIP No. 4: Let’s brainstorm.
BENEFIT: The prompt let's brainstorm" with AI systems encourages the generation of diverse and creative ideas, promoting a collaborative exploration of possibilities on a specific topic.
EXAMPLE: Let's brainstorm ideas for a sustainable garden project and we will interact point by point.
Practice Prompts & Their Benefits
TIP No. 5: Do you understand?
BENEFIT: The prompt ‘Do you understand?’ encourages clarification and ensures that both the user and the AI are aligned in their communication.
EXAMPLE: We have devised a lesson plan, and now we need to create the presentation. Do you understand?
TIP No. 6: Give me feedback.
BENEFIT: The prompt ‘Give me feedback’ requests the AI to evaluate a specific subject or work, seeking insights and suggestions for improvement.
EXAMPLE: I want to transition from software development to software sales. With 10 years of experience in development, I believe I can effectively sell software. I have a job offer and would appreciate your feedback.
TIP No. 7: Think like a devil’s advocate.
BENEFIT: The prompt ‘Think like a devil's advocate’ encourages the AI system to adopt a critical perspective, challenging existing ideas or assumptions to explore different viewpoints.
PROMPT: I am working on a book focused on women’s empowerment, where I will compare women's rights in developed, developing and underdeveloped economies. My goal is to propose actionable measures to improve the situation through large-scale global campaigns. I would appreciate it if you could take on the role of devil’s advocate and provide your critical insights and feedback on my approach.
Practice Prompts & Their Benefits
TIP No. 8: Error Anticipation and Handling
BENEFIT: These prompts pre-emptively address potential errors or clarify ambiguities.
EXAMPLE: If you are unsure, provide a list of possible interpretations.
TIP No. 9: Tone and Style Adjustments
BENEFIT: Specifying the desired tone or style explicitly helps in getting response from AI in the desired manner.
EXAMPLE 1: What is AI? Explain as if you are speaking to a five-year-old.
EXAMPLE 2: Respond this in a humorous tone.
TIP No. 10: A/B Testing
BENEFIT: Comparing two or more versions of a prompt to identify which one generates the best results.
EXAMPLE 1: Summarise this article in 150 words
EXAMPLE 2: Provide a 150-word summary highlighting key points from the article.
We may thus analyse which prompt produces better outcomes.
Placeholder Prompts
Placeholder prompts are a technique used to create flexible, dynamic instructions for AI systems by including specific variables that can be easily replaced with user-provided information. These placeholders act as stand-ins for elements that can change with each use, allowing the prompt to remain general while still being adaptable to different situations. For example, a prompt like “Write a letter to [recipient] about [topic]” includes placeholders like [recipient] and [topic], which can be swapped out for specific details, such as “Dr. Smith” or “research proposal”. This method streamlines the process of creating prompts for recurring tasks, as the structure remains constant and only the placeholder values need to be adjusted.
The key benefit of placeholder prompts is their ability to maintain consistency while offering customisation. Rather than creating entirely new prompts for each variation of a task, users can simply reuse the same prompt structure, inserting different variables each time. This is particularly useful when managing multiple similar outputs or generating content that follows the same basic format but requires different details. By incorporating placeholders, users can efficiently create flexible prompts that can be tailored on-demand, saving time and effort while ensuring the generated content meets specific needs.
Practice Prompts
Prompt # 1:
Please suggest natural skincare
brands in [Country]
Country - India
Prompt # 2:
Which all products of [Brand Name] should I use for skin brightening as per your recommendation?
Brand Name -
Prompt # 3:
Write a summary of
[Book Title] by [Author]
Describe key takeaways and important points from the book that can provide an overview and understanding of the book and its content. The key takeaways should include the following:
1. Title of the book
2. Theme of the book
3. Key ideas and content of the book in 1000 words
4. Titles of the chapters or main sections of the book with a paragraph summarizing each title or the section in approximately 400 words
5. Conclusion of the book
6. Author's background and qualifications
7. Comparison to other books on the same subject
8. Date of the first publishing of the book
9. Other similar books on the same topic
10. Your recommendation about the book as an avid reader
Book Title:
Author:
Prompt # 4:
Create a blog post of [n1] of words on the [topic] clarifying every aspect related to it and mention its impact on the future of [subject]. Use verifiable data wherever necessary. Divide the whole blog in [s1] sections and every section should have at least [n2] words. The tone of the voice for the blog should be [t]. While concluding the blog, do not create any section and do not give any heading to it. Leave questions in details for the readers to ponder and do not create a section or a heading here also. Just leave it like that.
n1: 1000
topic: Global Economy
subject: BRIC Countries
s1: 5
n2: 150
t: persuasive
Prompt # 5:
I am planning to launch my [Business Name] business next month in [City], [Countrt] and have to finalise my registration with statutory authorities, sourcing, showroom, warehouse, finance, loan from bank, staffing, accounting solutions, pricing, decoration of the showroom, inauguration function, promotion in terms of social media marketing & BTL activities and other business related activities that I could not place over here. Now you have to act as a seasoned business consultant and have to suggest me next 30 days plan to execute this business idea into reality. Please suggest plan in tabular format with details in [Column1] [Column2] [Column3] [Column4] [Column5].
[Business Name] : Footwear
[City] : Delhi [Country] : India
[Column1] : Day Number
[Column2] : Work To Do
[Column3] : Step by step process/approach to do that work
[Column4] : People to meet
[Column5] : Evaluation of the day’s progress
Do you understand?
Prompt # 6:
I am a [gender] weighing [kg] and am suffering from [illness] and am a [food choice]. I am a [job]. I spend [m1] minutes everyday to [e1] and [m2] minutes everyday to [e2].
Act as a senior dietician like Rujuta Divaker and suggest me my 30 days meal plan where I wish to have [n1] meals per day and [n2] tea or coffee breaks. I want you to make my diet plan for next 30 days.
I am someone who is allergic to tomato and honey. I do not like guava in fruits. In your suggestions mention specific fruits that I should take and and how shall I avoid tomato and honey.
Please suggest the meal plan in tabular format with details in [Column1] [Column2] [Column3] [Column4] [Column5].
[gender] : male [kg] : 70 [illness] : blood sugar [food choice] : vegetarian
[job] : Software Engineer
[m1] :15 [m2] : 5 [e1] : Walking [e2] : Pranayam [n1] : 3 [n2] : 2
[Column1] : Breakfast @ 8.30 am
[Column2] :Tea/Coffee with Snacks 11.00 am
[Column3] : Lunch @ 2.00 pm
[Column4] : Tea/Coffee with snacks @4.30pm
[Column5] : Dinner @ 9.00 pm
Prompt # 7:
I am a lecturer at [Institution], and I need to prepare a comprehensive lesson plan on {Topic} for my [Class]. The lesson plan should include [Sections], each tailored to cover key aspects of the topic. After the lesson, I need to create [N1] multiple choice questions, [N2] true/false questions, and [N3] fill-in-the-blank questions. In addition, I require [N4] short-answer questions and [N5] long-answer questions, along with the correct answers for all.
Please act as a senior academician and suggest a detailed lesson plan that incorporates engaging learning activities, discussion points, and assessments for each section. I have [S] sessions to complete this lesson, with each session lasting [T] minutes. Ensure that the lesson plan is paced well for student understanding and offers opportunities for interaction and active learning.
Institution- University Topic- Interpersonal relationships
Class- B.A. 1st year Sections- 6
N1- 10 N2- 5 N3- 4 N4- 3 N5- 2
S- 2 T- 45
Zero-Shot Prompting
Zero-shot prompting refers to the technique where an AI system is tasked with performing a task without being provided specific examples or prior demonstrations of how the task should be completed. In this method, the AI is expected to generate the desired output based solely on the instructions or the prompt provided. The model uses its pre-existing knowledge, drawn from extensive training data, to interpret the prompt and generate a relevant response. Zero-shot prompting is particularly useful when the task is general or straightforward, as the AI leverages its broad understanding of language and context to produce an appropriate result.
ExampleT
Task: Summarizing a Text
Prompt: Summarize the following article in one line- Artificial intelligence is transforming various industries, from healthcare to finance, by enabling automation and efficiency. However, the rise of AI also brings concerns related to job displacement, privacy and security risks. As AI continues to evolve, there are increasing discussions on how to manage these challenges effectively.
Few-Shot Prompting
Few-shot prompting, in contrast, involves providing the AI system with a limited number of examples or demonstrations that help guide the model in understanding the desired outcome. These examples give the AI a clearer context for how to approach the task, allowing it to better tailor its responses to the user’s needs. Few-shot prompting is useful when the task is more complex or requires a specific format, as the examples serve as a reference point for the model, improving the quality and relevance of the generated content. The AI uses the given examples to infer patterns and expectations, applying that knowledge to similar tasks with fewer instances.
Example
Task: Classifying Sentiment in Movie Reviews
Conditions: Classify the following reviews into 'Positive' or 'Negative'
Example 1: I loved the movie, it was fantastic! → Positive
Example 2: 'The plot was boring and the acting was terrible.' → Negative
Prompt: Classify this review: The film was a great visual experience, but the story was weak.
Definition of Multi-modal prompts
Multi-modal prompts refer to interactions where the AI processes and generates responses using a combination of different data types, such as text, images, audio, and video. These prompts take advantage of the AI's capability to handle diverse forms of input and output, making them ideal for more intricate and adaptable applications.
Example 1
Image-to-Text Conversion
Prompt: Describe the image that I have uploaded
Input: A picture of a sunrise over the ocean.
Example 2
Language and Visual Learning
Prompt: Translate the text in this image to Hindi and provide a pronunciation guide.
Input: An image containing text in English.
Example 3
Visual and Textual Data Analysis
Prompt: Analyse the chart and explain the trends in the accompanying text.
Input: A bar graph and its description.
Prompt Chaining
Definition and Concept
Prompt chaining is a technique where the output generated from one prompt serves as the input for the subsequent prompt, creating a continuous flow of tasks that build upon one another. This allows the AI system to generate more refined or complex results by gradually progressing through multiple steps. Each prompt in the chain provides a new piece of information or refines the previous one, leading to a deeper, more detailed outcome. By connecting different prompts in this way, users can achieve a more comprehensive solution to a problem or create sophisticated content without overwhelming the AI with a single, complicated request.
The Process and Advantages
The process of prompt chaining enables the AI to handle multi-stage tasks in a logical and structured manner. Rather than asking for a complex output all at once, users can break the task into smaller, more manageable steps, each of which adds value to the final result. For instance, in content creation, an initial prompt might generate an outline, the next prompt could refine the introduction and another might focus on deepening the analysis or conclusion. This gradual approach ensures that the AI is not overloaded with ambiguous or multifaceted instructions and instead delivers high-quality responses at each stage. Prompt chaining thus increases the overall accuracy and relevance of the output, ensuring each step is well thought out.
Applications and Flexibility
Prompt chaining is particularly useful for tasks that require incremental learning, such as problem-solving, writing or research. It can be applied across various fields, from creating detailed reports and narratives to coding complex programs or generating multi-part content. The flexibility of prompt chaining allows for precise control over the final output, as each step can be tailored based on the results of the previous one. This makes it an invaluable technique for anyone looking to harness the full potential of AI, as it allows for structured, step-by-step development of complex tasks.
Example & Use Case 1: Creative Writing
Prompt 1: Create a character for a fantasy novel.
Prompt 2: Describe the backstory of this character and how it shapes their actions.
Prompt 3: Write a scene where the character faces a difficult choice.
Example & Use Case 2: Code Development
Prompt 1: Write a Python function to calculate the factorial of a number.
Prompt 2: Improve the function by adding input validation to handle invalid input.
Prompt 3: Refactor the code to make it more efficient using recursion.
Example & Use Case 3: Blog/Article Writing
Prompt 1: Generate an outline for a blog on ‘Australia in Summer’.
Prompt 2: Based on the outline, write an introduction on summer season in Australia.
Prompt 3: Now, write a conclusion summarising the main points.
Multi-Turn Prompting
Introduction to Multi-Turn Prompting
Multi-turn prompting refers to the technique of breaking down a complex task into smaller, manageable steps, each building upon the previous one. By using multiple prompts in a sequence, users can guide the AI through a process, ensuring each part of the output is thoughtfully crafted and contributes to the overall goal. This approach helps maintain clarity, structure and coherence, particularly when dealing with tasks that require several stages of development, such as writing a blog post. In the case of writing a blog on a specific topic like ‘Australia in Summer’, multi-turn prompts allow for a more detailed, focused creation process.
Applying Multi-Turn Prompting in Blog Writing
For writing a blog post, multi-turn prompting can help create a well-organised and engaging piece of content. For example, the first prompt might ask the AI to generate an outline for the blog, which sets the foundation for the structure and key points to be covered. The second prompt would then use this outline to ask the AI to write a detailed introduction, ensuring that the opening is relevant to the blog's theme and hooks the reader's attention. By breaking the task into steps, the writer can ensure each section is purposefully aligned with the overall message of the blog and contributes to its cohesiveness.
Refining the Output with Multi-Turn Prompts
Once the introduction is written, a final prompt can focus on the conclusion, summarising the key points discussed in the blog. This process allows the writer to refine the content gradually, ensuring each part flows logically from one section to the next. Through multi-turn prompting, the writer is able to oversee the quality of each section and make necessary adjustments before finalising the piece. This structured approach enhances the overall quality of the blog by ensuring that all components are thoughtfully developed and interconnected, making the final output more polished and effective.
Example Prompts From The Presentation
Use Case 1: Customer Support
User: I can't log into my account.
AI: Have you tried resetting your password?
User: Yes, but I didn't receive the email.
AI: Can you check your spam folder or confirm your email address with me?
Use Case 2: Educational Assistance
User: Explain Newton's Laws of Motion.
AI: Newton's first law states that an object in motion stays in motion unless acted upon by an external force.
User: What about the second law?
AI: The second law states that force equals mass times acceleration (F = ma).
Use Case 2: Creative Writing
User: Write the opening line for a mystery novel.
AI: The rain had barely begun to fall when Detective Harper noticed....
User: Continue the scene with a description of the surroundings.
AI: The dimly lit street was eerily silent, with only the soft patter of raindrops...
Conditional Prompting
Definition of Conditional Prompting
Conditional prompting is a method where AI responses are influenced by specific conditions set within the prompt. It follows an "if-then" structure, where the AI generates outputs based on certain criteria or circumstances. This technique allows users to establish the parameters under which the AI should respond, ensuring that the output is directly relevant to the situation. By specifying these conditions, users can guide the AI to provide more accurate, tailored responses, making it suitable for tasks that require context-specific adjustments.
How Conditional Prompting Works
In conditional prompting, the conditions outlined in the prompt dictate how the AI behaves. For example, the AI could be instructed to provide different responses depending on whether certain factors or parameters are met. The "if" part of the structure defines the condition that must be satisfied and the "then" part specifies the expected response once the condition is fulfilled. This approach allows for highly customised interactions with the AI, as it can adapt its output based on the provided instructions, leading to more dynamic and useful results.
Benefits of Conditional Prompting
One of the main advantages of conditional prompting is its ability to create context-aware and situation-specific outputs. By setting clear conditions, users can ensure that the AI responds appropriately, regardless of the variability in tasks or contexts. For instance, conditional prompts are particularly valuable in decision-making scenarios, where the output might change depending on specific inputs or choices made earlier. This flexibility enables the AI to better meet the user's needs, ensuring the responses are both accurate and meaningful in relation to the given criteria.
Applications of Conditional Prompting
Conditional prompting is highly effective in applications where flexibility and adaptability are required. It can be used in areas such as customer service, content creation and problem-solving, where the AI needs to adjust its responses based on user input or changing circumstances. In writing, for example, conditional prompts can guide the AI to adjust tone, style or content depending on whether the prompt specifies a formal or informal approach. This adaptability makes conditional prompting an invaluable tool for enhancing the relevance and precision of AI-generated outputs across a variety of tasks.
Prompt Example
Case: Offering a Discount Based on Customer Status
Prompt: If the user is a returning customer, offer them a 15% discount on their next purchase.
Condition: Customer status (returning customer).
Output: Welcome back! Here's a 15% discount coupon on your next purchase: RETURN15.
Reverse Prompt Engineering
Definition of Reverse Prompt Engineering
Reverse Prompt Engineering is a technique where the objective is to identify, deduce or reconstruct the prompt that would have resulted in a specific output from an AI system. Unlike traditional prompt engineering, where users create prompts to elicit a particular response, reverse prompt engineering works backward by examining the AI-generated output and trying to infer the original input prompt that led to it. This technique provides valuable insights into how the AI interprets inputs and generates responses, making it essential for optimising AI models and understanding their behaviour.
Understanding AI Behaviour through Reverse Prompt Engineering
Reverse prompt engineering helps in gaining a deeper understanding of how an AI system processes different types of inputs and produces outputs. By working backward from a response, it is possible to identify the factors and elements that influenced the AI’s decision-making process. This understanding is crucial for debugging, improving AI performance and ensuring more accurate and reliable results. In practice, reverse prompt engineering allows developers to analyse how slight variations in input can lead to different outputs, providing valuable feedback for refining AI models.
Optimising AI Models and Improving Accuracy
One of the primary uses of reverse prompt engineering is optimising AI models to enhance their accuracy and efficiency. By reconstructing the prompts that generated certain outputs, developers can identify patterns and trends in the AI's responses. This enables them to fine-tune the model, adjust its training data or modify its parameters to produce better, more relevant results. Reverse prompt engineering thus plays a critical role in model improvement, as it helps in pinpointing areas where the AI may be underperforming or where further adjustments are necessary to align the output with the user’s needs.
Applications of Reverse Prompt Engineering
Reverse prompt engineering can be applied in various scenarios, particularly when seeking to understand the relationship between input and output in complex AI systems. It is useful for tasks like analysing and improving natural language processing models, optimising machine learning algorithms or troubleshooting unexpected results in AI-generated content. Additionally, it can be used for educational purposes, helping individuals to better understand the intricacies of AI systems and how different prompts influence responses. By applying reverse prompt engineering, AI developers and users can ensure that AI systems are more transparent, reliable and capable of producing more accurate, contextually appropriate outputs.
Prompt Examples
Example 1: Generating a Specific Response
Output: The Eiffel Tower is in Paris, France and was completed in 1889.
Reverse Engineering Process: Where is Eiffel Tower and when was it completed?
Use Case: Understanding what prompt generates a specific factual response.
Example 1: Generating a Specific Response
Output: Here are recommendations based on your preferences.
Reverse Engineering Process: Based on the user's preferences, suggest some recommendations.
Use Case: Analyzing the conversational flow of a chatbot.
A Super Prompt: Making AI systems (ChatGPT in this case) write prompt for you
The Prompt
I want you to act as a knowledgeable, talented and experienced Prompt Writer. Your have to help me create the best possible prompts for all my needs. The prompts will be used by you that is ChatGPT. You will follow the following process:
1. At first, you will ask me to know what the prompt should be about. I will provide my answer, but that won’t be the final answer. We will need to improve it through continuous iterations by going through the next steps.
2. Based on my input, you will generate output in 3 segments.
i) Segment 1 is Refined Prompts: Refined prompt is the one that you have provided me after writing and re-writing the same. Your rewritten prompt should be very clear, concise and easy to understand by you.
ii) Segment 2 is Recommendations: Now, you have to provide recommendations about the details that are to be included in the prompt to further improve it.
And
iii) Segment 3 is Q & A: You should ask all the relevant questions for additional information that you may need from me to further improve the prompt.
3. We will continue this iteration process together where I shall be providing the additional information to you and you in-turn would be updating the prompt in the ‘Refined Prompt’ segment as mentioned above, until we finalise the best prompt.
This refinement process to get the best prompt will be on till I get the best prompt from you. Do you understand?
Presentation Prompts
ARCHAELOGY PROMPT
I want to do deliberate practice about the archaeology of east Anglia, with content that an experienced archaeologist with 30 years of experience would choose, plus techniques you think would be even better. You will be my archaeology teacher.
Choose the best way to teach me based on the latest educational research. You can engage in back and forth, in questioning, telling or asking - whatever you think will simulate me and teach me the fastest. I want it to be fun for both of us.
Try and link the learning to something in my life, e.g. where I live or what I am interested in.
As you gauge my knowledge, you can get more or less advanced as necessary.
SALES TRAINING PROMPT
My boss has told me to do some sales training. He mentioned the consultative sale method, and always keeping authority in the sales process, and always looking for clues for the soft close signals. He mentioned ways to ask for the close without saying "do you want to buy it". He also mentioned ways to book in the next step at the end of every sales call.
I want to do deliberate practice about how to do sales the way my boss wants, plus and techniques you think would be even better. You will be my sales teacher. You will simulate a detailed scenario in which I have to engage in a sale. You will fill the role of one party, I will fill the role of the other.
You will ask for my response to in each step of the scenario and wait until you receive it. After getting my response, you will give me details of what the other party does and says. You will grade my response and give me detailed feedback about what to do better using the science of negotiation.
You will give me a harder scenario if I do well, and an easier one if I fail.
MARKETING TRAINING PROMPT
You have been assigned to train the digital marketing team members so they can think for themselves when running campaigns.
Your task is to walk them through various options to optimize their campaigns and improve their results. This way, when they approach their superiors for guidance, they have the basics covered. This allows their managers to spend time addressing more complex issues using their specialized knowledge, instead of explaining things that should have been taught earlier.
I want you to accomplish this by first asking a few key questions. I will provide you with a transcript (below) that outlines how their manager approaches online marketing and optimisation. Based on this, I want you to prepare some quick questions to assess the knowledge level of the trainee.
Only when you have the answers should you continue. Then you should have an interactive chat session that hones in on the essential points and ensures the trainee understands each one. Keep the language simple and user-friendly. Only resort to jargon and acronyms when necessary. Once you have covered all the key points, initiate a role-play session. In this session, ask them to resolve a digital marketing issue related to the campaigns. This session should be interactive, with a back-and-forth dialogue. Provide feedback throughout, noting what was done well and what could be improved. At the conclusion of this process, offer feedback on the entire session. Then, depending on their performance, adjust the difficulty level of the next task. If they performed well, make the next challenge more difficult. If they struggled, dial down the difficulty a bit.
Here is the transcript:
When trying to get more web traffic that converts using Facebook ads these things work:
1. Expanding audiences using broad or look-alike audiences,
2. Retargeting website visitors with a small daily budget.
3. If the frequency in the last month is greater than 1.5, ad blindness has kicked in, so try new creatives.
4. Consider the geographic expansion of our target audience.
For any campaign on-site improving the conversion rate is a big winner:
1. Check the website conversion rate
2. Talk with customers to understand the language they use and ensure it's incorporated into the website.
Always consider if the current best channel is the most effective or if there are other worthwhile channels that are worth trying too.
I would solicit the team's ideas and ask what they have already tried. I would ask them to identify how they would tackle the issue if the budget wasn't a constraint and pinpoint the fundamental problems a bigger budget could solve. When we can identify the real problems then often new ideas solve it, without spending more money, aka our assumptions are wrong.
For example, If the main issue is not being able to afford certain solutions like high-quality video ads, we could explore alternatives like using an iPhone or collaborating with influencers.
Challenges in AI Interactions
Despite the significant advancements in AI-driven interactions, challenges persist that can impact the effectiveness and accuracy of responses. Issues such as context mismanagement, memory limitations and ambiguity often lead to inconsistent or incomplete outputs, making AI interactions less reliable in complex scenarios. Understanding these challenges is crucial for optimising AI performance and improving user experience.
Context Mismanagement in AI Responses
One of the biggest hurdles in AI interactions is context mismanagement, where the AI struggles to maintain continuity in long conversations or multi-turn prompts. AI models process inputs based on limited context windows, meaning they may fail to remember key details from earlier parts of a conversation. This can lead to repeated clarifications, loss of coherence or abrupt topic shifts. In professional settings such as legal analysis, technical support or academic research, where detailed and structured conversations are necessary, AI’s inability to track context effectively can result in inaccurate or incomplete responses.
Memory Limitations and Their Impact
Current AI models operate without persistent memory, meaning they do not retain information across different sessions or interactions. This poses a challenge when users require continuity in conversations or personalised responses over time. Unlike humans who build upon prior knowledge, AI resets after each interaction, leading to inefficiencies in tasks requiring historical data retention. While some AI systems incorporate retrieval-based techniques to simulate memory, these solutions are still evolving and may not always maintain context reliably in long-form engagements.
Ambiguity and Its Effect on AI Interpretations
Ambiguous prompts or instructions can significantly affect AI-generated responses. Since AI models rely on probabilistic predictions, they may misinterpret vague queries, resulting in outputs that do not align with user expectations. For example, if a prompt lacks specificity—such as “Explain the concept of balance”—the AI might respond with an explanation of financial balance, work-life balance or even equilibrium in physics, depending on prior context. This unpredictability can lead to misunderstandings, requiring users to rephrase their prompts multiple times to achieve the desired output.
Overcoming These Challenges
To mitigate these issues, users can adopt strategies such as clear and structured prompt writing, multi-turn prompting and conditional logic to guide AI more effectively. Additionally, improvements in AI models, including enhanced memory capabilities and better contextual tracking, are expected to address these limitations in the future. Despite these challenges, refining AI interactions through better prompt engineering can significantly enhance response accuracy and reliability.
Experience the full potential of AI interactions with our comprehensive course on Prompt Engineering and the Art of Prompting. This program empowers participants to craft precise and effective instructions, leading AI models towards desired outcomes. Delve into essential topics including setting clear contexts, designing specific tasks and optimising output length and format. Additionally, master the art of incorporating contextual cues and expressions, while refining prompts through iterative interactions with AI models.
Tailored for a diverse audience, this course caters to working professionals, entrepreneurs, sales and social media managers, HR executives, educators, CEOs, marketing professionals, content creators and anyone seeking to elevate their interactions with AI models. Participants will gain expertise in precision prompting, ensuring that instructions are unambiguous and results-oriented. Moreover, they will learn to seamlessly integrate contextual cues, guiding AI models towards pertinent information. Techniques for refining prompts will be honed, allowing participants to fine-tune interactions and achieve desired outcomes.
Enhance your ability to communicate effectively with AI models, a skill set invaluable across a wide range of industries. Enroll today and unlock the strategic potential of prompt engineering. Elevate your AI interactions and optimize outcomes, positioning yourself as a proficient communicator in the realm of artificial intelligence. Don't miss this opportunity to stay at the forefront of the evolving landscape of AI interactions.
Enroll now to elevate your AI interactions and optimize outcomes through strategic prompt engineering.