Thanks to AI-based tools like ChatGPT, the use of generative AI has exploded in 2023. In fact, the rise of generative AI was easily the top tech story of the year. As such, I thought that it might be fun to reflect on how generative AI has affected my work.
Before discussing how I use generative AI, I would like to stress an important point. We all know that generative AI tools such as ChatGPT and Microsoft 365 Copilot can produce written articles, blog posts, and other types of content. Let me say upfront that I absolutely do not use generative AI to write content. While it is true that I could use ChatGPT to create in seconds something that would ordinarily take me a couple of hours to write, I choose not to do that. After all, if sites like this one wanted AI-generated content, they could easily create it themselves.
The advantage of human-written content is that humans learn by doing rather than by mindlessly regurgitating information from the internet. During my over 30 years in IT, I have gained all kinds of tricks, techniques, and experiences that I try to incorporate into my writing. Generative AI lacks that sort of real-world experience.
So, if I don’t use generative AI for content creation, you may be wondering what I do use it for. Here are the different ways that I use generative AI on a daily basis.
As a Research Tool
I primarily use generative AI to conduct research, as an alternative to a traditional search engine. This habit originated from an experience that had nothing to do with work.
Last spring, my wife and I were trying to plan a vacation but just could not decide on a destination. Since we were getting nowhere with our trip planning, I turned to ChatGPT for help. While I didn’t expect ChatGPT to find anything different from our searches, it surprised me by quickly suggesting a perfect vacation spot with very little effort on my part.
I quickly realized that ChatGPT had done what a search engine couldn’t. When you enter a query into a search engine, you receive a list of web pages that may or may not contain the information you seek. It’s up to you to navigate the pages to search for information. ChatGPT, meanwhile, provides a direct answer to your question by aggregating information from various websites.
I have found this ChatGPT capability can save me a lot of time when I research an article, as I don’t have to sift through a million web pages.
However, it’s important to recognize that generative AI tools, including ChatGPT, can occasionally provide incorrect answers. That means you should validate the information you receive. For instance, if I ask ChatGPT how many virtual processors a Hyper-V virtual machine supports, I could validate the answer by asking where the information came from. From there, I could check out the source of the information for myself. Another thing I could do is perform a Google search to cross-check and verify the accuracy of the information ChatGPT gave me.
As a Learning Tool
As a freelance author, I can sometimes find myself writing about subjects I know little about initially. After all, no one (me included) can know everything. That being the case, I have occasionally used ChatGPT as a learning tool.
In these situations, I ask ChatGPT to explain the subject in an easy-to-understand manner. The advantage of this approach is that, after reading the response, I can ask ChatGPT follow-up questions. For example, if I wanted to learn about the concept of orbital decay, I could prompt ChatGPT to provide a simplified explanation, bypassing all the crazy mathematical calculations and focusing on basic aspects of atmospheric drag. I might then ask further questions, such as whether factors such as tidal conditions or solar winds contribute to orbital decay.
As an Outlining Tool
Occasionally, I also use generative AI as a tool for figuring out what I should include in an article I’m writing. The publications I write for often will provide only vague descriptions of the topics they want me to cover. For example, I might receive an email that says something like, “Can you write something about Hyper-V checkpoints?”
There are millions of ways that I could approach an article on Hyper-V checkpoints. I might write about checkpoint creation, performance considerations, broken checkpoint chains, or even checkpoint management via PowerShell. In these situations, I may ask ChatGPT about the questions people most frequently ask about the topic. This gives me some insight so I can then address the most relevant and engaging aspects of the subject for readers.
In this video tutorial, Brien Posey explains how to install and use Copilot in PowerShell.
As a ‘Getting Started’ Tool
It’s strange to consider that, even though I have been writing professionally since 1995, I can still find myself staring at a blank page, wondering where to begin. Once in a while, I turn to ChatGPT as a tool to help overcome writer’s block.
For example, suppose that I am tasked with writing a five-page article on Windows IIS. While I know basically what I want to say, I may still have trouble kick-starting the writing process. I might ask ChatGPT to write a two-paragraph article on the same topic. Reading the first sentence or two shows me how ChatGPT introduced the topic. While I make it a point to avoid mimicking ChatGPT, the sample article will often give me ideas about how to get started on my own article.
About the author
Brien Posey is a bestselling technology author, speaker, and 21x Microsoft MVP. In addition to his ongoing work in IT, Posey has trained as a commercial astronaut candidate in preparation to fly on a mission to study polar mesospheric clouds from space.
