Week 5 - BALT 4362 - Prompt Engineering

-

 Before I get into the prompt, it is important to define what prompt engineering is. Prompt engineering is the "skill" or "art" of asking specific, defined questions or giving clear instructions to AI so you can get the response you want. AI can only be as useful as the prompt it is given. With that being said, this is the prompt I gave ChatGPT: "Suggest a positive but firm response for a customer who is unhappy with their bill. [Paragraph]" 

This is the response ChatGPT gave: "I completely understand your concern and appreciate you reaching out about your bill. We strive to ensure all charges are clear and accurate, and I want to reassure you that I will personally review your account to confirm everything is correct. If any adjustments are needed, we’ll take care of them right away. Thank you for bringing this to our attention—we value your business and are committed to making sure you feel confident in the service you’re receiving." 

This is a nice response, but it was not exactly what I was thinking. I was trying to generate a response in a scenario where the customer came to the business to argue their bill. This response would work for a phone call or an email exchange. The solution is to refine the prompt even more. I then said that the response was good, but it needs to be conversational and not so straightforward. 

"I hear your concern, and I understand why you’d be frustrated. What I can do right now is take a closer look at your bill with you so we can make sure everything is correct. If there’s an error, we’ll fix it, and if the charges are accurate, I’ll walk you through them so they make sense. Either way, I want you to leave here feeling clear and confident about what you’re being billed for."

This response is more of what I was looking for. It is firm, reassuring, and makes the customer feel like they are being listened to.



Comments

Post a Comment

Popular posts from this blog

Jude's Introduction Post 4361-4364

-
Image
Welcome to my blog!  My name is Jude Sullivan, and this will be my third Fall here at Benedictine. I am originally from Paris, Illinois but made the decision to move across the country and continue my baseball career in warmer weather, at the Mesa campus. I played on the baseball team my first two years, but I have decided to focus on school/work for this Fall semester. I will be graduating in December with an Accounting degree, and I am looking forward to learning the skills that this class will require.
Read more

Week 8 - BALT 4362 - AI Image and Video Generation

-
Image
AI’s ability to create realistic images and artwork has exploded in the last few years, and in my opinion, it is both amazing and terrifying. With the correct prompt, anyone can make professional-quality pictures, no art degree required. Businesses can now use tools like DALL·E, Midjourney, and Canva’s AI generator to design posters and logos in seconds. This levels the playing field for small businesses that do not have the marketing budget to compete with bigger corporations. This can also be applied to students. You can make visuals for class projects, creative writing, or social media without needing advanced design skills. I know I have been able to use it to create presentation visuals I never would have been able to come up with on my own. It does a great job bridging the gap for people who might lack creativity, like me.  But as cool as it is, there are some real downsides. Artists worry about AI copying their work or taking credit for their style, and some AI-generated im...
Read more

Week 2 - BALT 4363 - CH 2

-
Image
Chapter Two of Data Toolkit,  and more specifically the exercises within the chapter,   helped me understand the basics of how Python can control and organize what a program does. Control structures like loops are like giving a set of instructions that the computer can follow repeatedly or only under certain conditions. The examples using “for” and “while” loops made it clearer how repetition works. An example of this would be repeating something a certain number of times or until a condition changes. I liked the comparison to a remote-control car, because it helped me see how code can “steer” what happens in a program. Seeing the examples with “if” and “if-else” statements helped me see how Python can make decisions, like checking whether a number is even or odd. The guessing game example at the end was especially interesting because it combined all the ideas, loops, conditions, and user input into one project. It made Python feel a little less intimidating, since I could se...
Read more