We always feel that we are wasting our time and draining our energy dealing with artificial intelligence, and we feel disgusted and angry at the stupid answers the AI gives us. I have a very bad experience with GPT; I spent a month trying to reach a goal, and it ended up with me deleting the app! However, by implementing these six steps, you will not need to modify the prompt every time or copy ready-made prompts, and your dealing with artificial intelligence will be easier, faster, and more effective.
Defining the Identity of Artificial Intelligence: Who is it?
Giving a role and identity to artificial intelligence saves you a lot of time, so that you confine all the AI's energy and focus to the role you gave it.
Suppose a negligent student wants to understand a specific lesson; we write to it, "You are now an expert and a professional teacher, and you can simplify information for beginners." Here, you have defined its exact role, and it will work according to the specified role.
Requirements: What does it do?
We define the requirements clearly and accurately to achieve the greatest benefit. Artificial intelligence loves details and clearly defining requirements.
Continuing the student's example, we write to it, "In the subject of chemistry, lesson X, explain it to me in detail, I am a beginner." As shown, we specified a particular lesson; if we tell it to explain something general like the science of chemistry, it will give you a general result as well.
The Purpose of the Request: For whom?
The more details we mention, the more wonderful the result will be. You might ask yourself, should I mention to artificial intelligence the desired goal of the request!? Yes, because it loves details.
Continuing the negligent student's example, we write to it, "I want you to explain lesson X to me because I have an exam tomorrow." There is a difference between telling it I have an exam tomorrow and leaving the prompt general without mentioning the purpose of the request. After you write to it that you have an exam tomorrow, it will try to explain the lesson to you quickly because your time is tight.
Imposing Constraints: What is the thing you do not want it to do?
Here we alert artificial intelligence to the prohibitions it must avoid, so that previously we used to burden ourselves and write 10 prompts until we reached what we wanted, but by mentioning the prohibitions and constraints here, we do not need to modify the prompt 10 times.
Continuing the student's example, we write to it, "Do not mention complex terms to me, do not elaborate excessively in the explanation and stuff useless words, do not explain in colloquial dialect." Here we saved ourselves time and we will get an excellent result.
Give it an Example: What is the final form of your request?
The final touches to the prompt: after we defined our goal and imposed constraints and prohibitions on it, what remains for us is to give it an example of the final form of the required outputs.
Continuing the student's example, we write to it, "Explain lesson X to me in the following way: you write the chemical equation first, then explain below it each element on its own, and how the elements interact with each other." With this touch, we have finished the fifth step, and here we can press send, send the prompt, and get the required outputs with extreme accuracy, but to make the matter easier in the future, we implement step number 6.
Pinning the Prompt in the AI's Memory:
We all overlook this step even though I consider it one of the most important steps that saves time; simply, we give artificial intelligence a two-letter code or a number, choose whatever you like.
Continuing the negligent student's example, after we implement the five steps, we write to it, "Save this prompt and pin it in your memory with the two letters NR; when I write to you only NR, you execute this prompt." In this way, we have spared ourselves the trouble of writing this long prompt again.
Conclusion
In the end, artificial intelligence loves clear details and loves for you to break tasks down for it; the more accurately and clearly the prompt is defined, the more accurate the outputs come to you.