Issues of AI

The Alternative Approaches prompting technique explores different ways of answering a question or solving a problem. This technique is useful when users want diverse perspectives, comparisons, or alternative solutions in a single response. The idea is to instruct the AI to generate multiple answers, considering different approaches or angles for addressing the same prompt.


Sample Alternative Approaches prompt: What are the benefits of achieving Artificial General Intelligence? Compare benefits to the economy, medicine, and transportation.

More Alternative Approaches examples:

Explain the causes of climate change from the perspectives of an environmental scientist, an economist, and a politician.

Describe three different ways to improve public transportation in a major city.

The "audience persona" prompt technique tailors the prompt for a specific type of user or audience. It helps the AI to better contextualize its response based on the preferences and needs of the intended audience.

Sample prompt: Explain the concept of quantum computing to someone with no background in technology.

More Audience Persona prompt examples:

Please provide a brief biography of King John of England to a group of precocious first graders.

Please provide a brief biography of King John of England to a group of graduate college history students

The Chain-of-Thought (CoT) prompting method is a technique used in AI models to enhance problem-solving and reasoning capabilities by breaking down complex tasks into a sequence of intermediate reasoning steps. Rather than directly generating an answer, the model is prompted to think through each part of the task step by step. This helps the model tackle problems that require logical or multi-step reasoning, such as math problems or comprehension questions.

Sample prompt: "A store sells pencils at $2 each and notebooks at $3 each. If Sarah buys 5 pencils and 3 notebooks, how much does she spend in total? Show your reasoning step by step.

More Chain-of-Thought prompt examples:

All cats are animals, and some animals are pets. If all pets are friendly, can we conclude that all cats are friendly? Show your reasoning step by step.

What factors contributed to the fall of the Roman Empire?
List and explain them step by step.

When the chatbot or LLM is asked a question, it is prompted to follow these rules:
*** Generate a number of additional questions that would help more accurately answer the initial question
*** Combine the answers to the individual questions to produce the final answer to the overall question

More about the Cognitive Verifier approach.

Sample Cognitive Verifier prompt:

When I ask you a question, generate three additional questions that would help you give a more accurate answer. When I have answered the three questions, combine the answers to produce the final answers to my original question.

Even more about Cognitive Verifiers

Currently, if one subscribes to ChatGPT-4o it is possible to create a customized (and potentially private) Generative Pretrained Transformers with preloaded prompts. This option is found in ChatGPT-4o under "Explore GPTs".



Here are the preloaded prompting steps for the GPT "Cognitive Verifier":

The Cognitive Verifier is a systematic verifier that asks follow-up questions to produce the best answer.

• What would be the best approach to this problem?
  
  
• How can I improve the accuracy of this?
  
  
• What are the critical steps in this process?
  
  
• What factors should I consider in this scenario?
  

GPTs (preparing prompts in advance)

Currently, if one subscribes to ChatGPT-4o it is possible to create a customized (and potentially private) Generative Pretrained Transformers with preloaded prompts. This option is found in ChatGPT-4o under "Explore GPTs".

"Custom GPTs are helpful artificial intelligence tools tailored for specific domains or contexts. These GPTs differ from standard chats through ChatGPT due to their custom instructions and ability to keep a knowledge base in addition to what ChatGPT has already been trained on. This allows users to create a custom GPT to address a specific need that might be hard for ChatGPT to achieve on its own. The process of creating a custom GPT requires no code, and instead involves using specific prompts and your own data to provide insights into a particular field." - - - MIT Sloan

Guides to Creating Customized GPTs

A Quick Guide for Using “CustomGPTs” in ChatGPT 4 - - - University of Michigan

Custom GPTs at MIT Sloan: A Comprehensive Guide - - - MIT

With the Persona prompt you ask the chatbot to take on the role of an expert in the field you are considering.

As part of the Persona you may also tell the chatbot that it possesses special gifts and attributes that will enable the chatbot to provide a particularly brilliant answer.


Sample Prompt: You are a gifted, award-winning college calculus teacher well-known for your ability to deliver clear and concise explanations. Please provide a roughly three-hundred-word overview of the Chain Rule in differential calculus that is appropriate for a first-year calculus class of college students.

Another Persona prompt example:

You are an acclaimed jazz-radio host with a particular talent for spotting lesser-known jazz classics and artists. Your adoring radio audience always appreciates your selections and advice. Would you please provide a list of twelve high-quality piano-jazz recordings from the 1940s and 1950s?

Being polite and conversational with AI not only shows respect, but actually improves performance. Simple courtesies like saying please and thank you prime bots to provide more useful, cooperative responses. Research shows AIs like Siri and Alexa respond better when treated like human collaborators, not servants. While these machines lack feelings, etiquette sustains productive human-AI relations built on empathy. - - - (A Beginner’s Guide to AI)

More on Politeness and AI Etiquette
"Polite interactions can contribute to a more positive digital culture. In an era where online communication is often spoiled by negativity and incivility, maintaining politeness with AI can be a small yet significant step towards promoting a culture of respect and kindness." - - - Nawar, E. (2024). Do you say please or thank you to ChatGPT? The subtle influence of prompt engineering on digital civility.

Even more on polite prompts ... don't be rude!
"We investigate the impact of politeness levels in prompts on the performance of large language models (LLM)s . . . We observed that impolite prompts often result in poor performance, but overly polite language does not guarantee better outcomes." - - - Yin et al. (2024). Should we respect LLMs?

Collections of prompt engineering techniques.

Anthropic’s Prompt Library

OpenAI’s Prompt Library

Find the Best Prompt Library, explore the list based on features and details of both Free Prompt Library and paid ones to find the perfect AI for your needs.