Practicing Perplexing Prompts

05/31/2024 10:38 AM | Paul Venderley (Administrator)

One of the participants in last night’s Practicing Prompts session asked about Perplexity.ai.  We’d been so focused on using our existing source material to create the structure of the course that we didn’t get into Perplexity’s wheelhouse.  Let’s explore that (briefly) here:

Perplexity.ai is designed for gathering the information that we can use for other content, such as our Writing Proposals course.  For example: we have an objective: “Incorporate relevant graphics and charts into the proposal to enhance understanding and engagement.”

A clipart of a telescope. Readers may skip words, but they will not skip images.

When I review my source material, I find a section that states: “Graphics are a powerful way to convey ideas and focus the reader’s attention.”  From there, it discusses selecting clipart or photos that show “some creative association with the associated text” such as a clipart image of a telescope as a metaphor for focus.

Bearing in mind that this source material was written in 2002.

I believe that this section could benefit with some more thought. Let me turn to Perplexity to see if there are any studies or recent references that will help us better relate how to include graphics into our proposals.

Where we may use a framework to design our prompts in ChatGPT or Gemini, I’m more comfortable asking Perplexity a starting question.  Perplexity will prompt me to ask a follow-up question if I’m not finding the resources I want.  I will state the role and use my source content to help structure my question.


Here’s where Perplexity shines:  It provides ideas to explore, and links to the content where it found those ideas, thus providing me with a starting point and a path for further exploration of the topic.

Right below the question, you can see a grouping of five sources. Select one of those boxes, you'll be directed to the site in question.  

Review Perplexity's answers, and you will see links to the article that the system is referencing.  For example, I'm able to open the "openasset" article to learn more about their ideas for Representing Data Visually.  That's not a bad article (despite their application of the 80/20/10 myth*) and one that did not appear in the results when I conducted a Google Search for "Selecting Proposal Graphics."

In essence, Perplexity is more of a search engine than a generative tool, albeit one that is capable of composing conclusions based upon the information it finds.  I tend to appreciate its guidance; it is a great resource for helping me identify what I don’t know and pointing me down the path to learn more.

Case in point.


*Ask Perplexity.ai: "Is it true that we remember 80% of what we see, 20% of what we read, and 10% of what we hear? What studies show these statistics?"  Share what you find in the comments.





Sponsors and Partners




Powered by Wild Apricot Membership Software