Prompt Engineering Tips

Have you ever been on the receiving end of vague directions? Like being told to “turn left after the big tree” only to be surrounded multiple big trees. It’s frustrating, right?

And, if you’ve spent any time playing with AI, I’ll bet you’ve felt that all familiar frustration there too; you’ve asked what you think is a simple enough request only to get a load of crap back, which makes you wonder how AI can save you time if you have to keep going backwards and forwards.

The thing to remember with any system is, if you put crap in, you’ll get crap out.

When it comes to AI, prompt engineering is the key to unlocking it’s potential, and while it might sound very complicated it’s simply learning how to craft precise and effective inputs or “prompts” to guide AI systems to produce the desired output. The clearer and more detailed your prompt, the better the AI can understand and deliver what you’re looking for.

 

Tips, Tricks & Frameworks

Let’s face it – time is precious so rather than leaving you to figure it out through trial and error here’s a few tips you can use to help you get what you want…

Short vs. Long: Sometimes a crisp, concise command works wonders, other times you need to paint a broader picture with a bit of context.
Cut The Waffle: Be direct and to the point. Try and channel that stern, but fair, gym instructor.
Use Open-Ended Questions: These invite the AI’s creativity which might bring you results that you hadn’t considered.

You can also explore using frameworks to structure your prompts more effectively:

 
Situation, Task, Action, Result (STAR):
  • Situation: Set the context for the AI.
  • Task: Define the task you want the AI to help with.
  • Action: Specify any actions or constraints.
  • Result: Describe the desired outcome.


Example: “You’re a marketing expert (Situation) tasked with creating a launch email for a new online course (Task). Use a friendly and approachable tone (Action) to entice potential students to sign up (Result).”

 

Who, What, When, Where, Why, How:

Covering the basics can prompt AI to fill in gaps creatively.

Example: “Who: Female entrepreneurs. What: Need promotion strategies. When: Next two weeks. Where: Online platforms. Why: Increase course enrolment. How: With minimal budget.”

 

Problem, Purpose, Perspective:
  • Problem: State the issue or challenge.
  • Purpose: Clarify the goal.
  • Perspective: Specify any unique perspectives or tones.


Example: “Problem: Course registrants aren’t converting into active participants. Purpose: Increase engagement by 20%. Perspective: Use a motivational and supportive tone.”


Role-play Scenario:

Create a fictional role for AI to assume, making it easier for AI to follow instructions.

Example: “Imagine you’re a productivity coach helping a busy mom start her online business. Suggest five time-saving strategies that blend work-life balance principles with business growth.”

 

So, go and have a play!

Remember, the more you practice, the better you’ll get at helping AI to serve you. Adopting a trial-and-error mindset is important, little tweaks and refinement can make a huge difference.

Let me know how you get on and if you’re ready to take it up a notch by creating your own AI Assistants, then drop me a message and let me know what you’d like your assistants to do for you and I’ll help you to refine your instructions.

Now, go forth and prompt like a pro!

If you found this useful, please do let me know, and if you’ve got any questions that you’d like me to cover in future episodes, or topics that you’d like to learn more about, either drop me an email, or come and join the conversation on my socials