ChatGPT Part 7

Adding Tonality To The AI’s Output

John Iovine



I want to explore a little further how to direct the AI to create work with a specific tonality. We can ask the AI to make the generated text that is: funny, sarcastic, diplomatic, helpful, loving, hateful, etc.

I already posted examples of directing the AI to assume the personality of someone. You can double down on the prompt by adding a personality that enhances the desired tonality. For example, you want a sarcastic joke about an obese person ordering a second helping of chocolate cake. Direct the AI to write the joke as the comedian Don Rickles.

Make it Funny

Please write a sarcastic joke about an obese person ordering a second piece of chocolate cake in the voice of comedian Don Rickles.

Whether you like to joke or not is not the point. The AI, in its response, is providing you with fertile ideas and concepts you can use to jump off from. You can choose to use only the best bits of the AI’s output. For instance, suppose you have a character, and you want him to say something to ride another character who ordered a second piece of pie during their meeting in a diner. The character could say, “Hey, is the baker giving you a group discount?”

Make it Sad

Write 50–75 compassionate and sad words that a father says to his son, telling him his pet cat needs to be put to sleep by the veterinarian.

Make it Angry

Write 50–75 angry words a man says when he is two minutes late returning to his car and discovers a parking ticket already issued by the police.

Make it Diplomatic

Write 50–75 diplomatic words a doctor says to an obese patient to encourage him to…



John Iovine

Science writer, thinker, self-experimenter, focusing on personal development and health —