Think of ChatGPT as a friend
As interested as I am in Artificial Intelligence (AI), and as pervasive as it is becoming in our daily lives, I seem to only manage one blog post per year about it. Suffice it to say that a lot can, and does, happen in a year! So, consider this your annual update.
Progress!
Click image to read “An Interview with ChatGPT” (2023)
For example, some of you will remember what ChatGPT had to say, in my post, An “Interview” with ChatGPT (April 2023), under “Reality Check”, about the books I have written. It named three books with entirely plausible-sounding but utterly fabricated titles. Not only had I not written them, but no one else had either!
At that time ChatGPT had no direct access to the internet and its knowledge cutoff date was September 2021. This time, however, it gave me the correct titles of my three books PLUS clickable links to their pages on Amazon and Goodreads!
Why am I writing about AI?
I’m no AI expert. My goal in writing about it is not to convince you to use it. I just want to demystify it a bit and let you know how I’ve been using it. One of my organizing colleagues, Jamie Steele, a productivity coach at Tidy Your Time, LLC, refers to ChatGPT as “Chatty G” which I think is adorable. It also dovetails with a thought I’ve had that might help you, too, when you are using it: Think of ChatGPT as a friend. (Keep reading to see why.)
Contents:
- AI affects us all
- Think of ChatGPT as a friend
- Some of the ways I’ve used ChatGPT lately
- CAUTION: Things to watch out for
AI affects us all
Examples of AI in Everyday Life. I know you can’t read this. Click the image to see the search results and links to more info.
AI is kind of like politics: It affects your life whether or not you are interested or pay attention to it. Here are some Examples of AI in Everyday Life (Click the link or the image to see the larger, clickable search results)
In The AI Gold Rush (May 2024) I mentioned that there were now thousands of new AI applications. They mainly fall into the two following categories:
- Predictive AI such as Google and Netflix searches
- Generative AI such as chatbots, image creation, research and writing tools
It’s more complicated than that, of course, but notice that chatbots (like ChatGPT) are not the same as search engines. They are Large Language Models (LLMs) with increasing access to the internet, but they do not replace search engines. If you want to know something factual, especially if it happened yesterday, it’s best to Google it. There are many different chatbots available, as discussed in this representative article: The best AI chatbots in 2025 (Miguel Rebelo for Zapier, March 4, 2025)
Click image to read “The AI Gold Rush” (2024)
If you aren’t sure which one to pick, there’s no reason not to try them all. Most offer a free version and a paid version with extra features. Some allow a certain amount of usage per day and that provides a built-in opportunity to try another one (if you really have that many questions or uses for it). You can even install Chrome’s Multi AI Sidebar to make it easy.
For purposes of this blog post, though, I’m sticking with free ChatGPT 4o. I think my head might explode if I branch out much further!
Think of ChatGPT as a friend
What do I mean by that? Well, I take everything Chatty G says with a grain of salt, just like I would if I were discussing a problem with a friend. Friends are (often) not doctors or lawyers or historians. Friends can make mistakes. But that doesn’t mean it’s worthless to ask a friend for advice or run something by them. Some friends are more knowledgeable than others, and sometimes they confirm things you already think, or steer you away from faulty thinking, and add some insight you’ve never considered. They can help you organize your thoughts, or your data.
I have heard that some people use chatbots as therapists of sorts. I mean, Chatty G is not a licensed healthcare professional, but it’s a good sounding board. If it’s not a life-or-death issue, why not run it by Chatty G? It might even advise you to seek the help of a therapist!
Some of the ways I’ve used ChatGPT lately:
Personal Memories
No amount of Googling was helping me find the name of the festival I used to play in each year as a piano student in 1970s Boise, Idaho. It didn’t really help that I was calling it a “competition” because it wasn’t, really. There were judges and ratings, but there were as many high (or low) ratings as there were students who deserved them. One of my memories was that I earned six Superiors in consecutive years and was therefore awarded two “Gold Superiors” (one for each set of three years). Well, they’re apparently called “Gold Cups” these days (trophies instead of certificates), and after I provided a few more details ChatGPT figured it out. (I was going to share the entire chat, but the link isn’t “sticking” for some reason. Contact me if you are curious about how the conversation unfolded, and what the festival was like.)
Genealogy
AI is used for all sorts of genealogy tasks, like record transcription, image enhancement, and language translation. All of the big platforms like Ancestry, FamilySearch, and MyHeritage are constantly leapfrogging each other in terms of tools and capabilities. I’m still following the Facebook group called “Genealogy and Artificial Intelligence (AI)” where even the genealogy experts are still experimenting and sharing both their successes and their failures. Most of them have either developed, or will soon, their own instructional videos, blog posts, and classes related to AI usage. And, for those not necessarily into genealogy, Blaine Bettinger has just created a brand new Facebook group called “Generative Artificial Intelligence (AI) Tools & Tactics”.
I used Stephen Little’s prompt (“Fun Prompt Friday!” posted on Feb 28, 2025) to create a “Parke County, Indiana Locality Guide”. Having gathered most of the sources myself over the years I can report that it did a pretty good job!
I also fed ChatGPT a family history I once wrote to see what it might notice about it. I asked questions like this:
- Are there any conclusions you can make that are not already stated in the article?
- Do you have any suggested resources for further research into the lives of these families?
I can’t say I’ve broken my brick wall on that branch of my family yet, but ChatGPT did give me some historical context and other things to think about and try!
In An “Interview” with ChatGPT (April 2023), under “Should we fear AI?”, I wrote about AI-generated images. Take a look at this video to learn more about how AI can affect our use of historical photos: Fact or Fiction? Comparing Photo Tools with ChatGPT. (I will admit to only having watched the first 6-1/2 minutes of it. So you have my permission to stop at that point, too.)
Mosquitoes
Chances are that mosquitoes are more of an issue for me than they are for you. For one thing, my blood type is O negative, which has been shown to attract more mosquitoes than other blood types. Second, we have a new stealth breed of mosquitoes here in New Mexico — tiny, silent, goes for the ankles, invades homes (well, my home, anyway). Third, I have lymphedema, and scratching mosquito bites can quickly lead to infection.
I had quite the chat with Chatty G about mosquitoes, remedies, and prevention methods. (Not sure why the link doesn’t “stick”, so contact me if you’d like to see the chat in its entirety. I may even write a whole new blog post about mosquitoes next.)
Writing
I have yet to really use AI for writing. But I know a lot of writers with questions about AI. I never even used grammar- or spell-checkers until they became default features of word processing programs such as Co-Pilot for Word. I have, however, like most writers who have ever lived, used a human editor, a thesaurus, and other such writing tools. I also Google words to make sure they mean what I think they mean, and topics to learn more about them.
But I do want to share that there are uses for AI that won’t get your term paper sniffed out by AI detectors or your book rejected by Kindle Direct Publishing (KDP). The KDP link describes the difference between material that is AI-generated (requires disclosure) and AI-assisted (does not require disclosure).
Surprisingly, I found this post by Open AI, the creator of ChatGPT — Writing with AI: Five ways professional writers are leveraging ChatGPT — to be more helpful than self-serving. Among other things, it says, “Writers are using ChatGPT as a sounding board, story consultant, research assistant, and editor—to talk out ideas, find the right word, clarify their thinking, and get feedback on structure and flow.” Another good one along these lines is this: Does Using Grammarly Make My Content get Detected as AI-Generated? The Results Are In (Jonathan Gillham for Originality.ai, October 24, 2024) Spoiler alert: It depends.
One thing to watch for in reading articles these days is whole chunks of text, or distinctive phrases, that are repeated. Bad editing? Or AI? Sometimes it’s both.
Medical
In my experience doctors often fail to sufficiently warn patients of drug interactions and ways to make them more or less effective. Or, even if they do, it’s easy for a patient (OK, for me) to forget. And, if it’s a medication you’ve taken for years, you may not think to ask a pharmacist about it and may still not know (or have forgotten) it needs to be taken on an empty stomach, or that you should avoid dairy, or citrus, or sunlight, or whatever, while you are taking it. Google has a lot of helpful info on drugs but sometimes it’s hard to assemble it all in a meaningful manner.
I had a couple of new medications that I wanted to ask ChatGPT about. I think I’ll keep this chat to myself, but the bottom line was that I did change my medication schedule as a result, but not precisely the way ChatGPT outlined it for me.
Voice Mode
The other day I was trying to work with Ziggy, my cat, in my arms. So I clicked the microphone on, thinking it was a dictation feature like I have in Word. I was expecting to see my spoken words typed out, and for Chatty G to type an answer in return. But no. Instead, I was offered several different Chat personas, with assigned names and unique voices. I chose one and asked “her” something about Social Security (without mentioning my actual SSN, of course). I noticed that chat got auto-labeled “Email clarification request” (which is not wrong, but which wasn’t helpful in searching for it later, whereupon I edited the chat name to something more meaningful). I noticed there were two voice-looking modes and asked her:
Q: “What is the difference between dictation and voice mode?”
A: “Dictation mode is usually for converting spoken words into text, like when you’re using speech-to-text. Voice mode, on the other hand, involves interacting with a device or software through spoken commands, like having a conversation.”
OK, got it! No wonder people like chatting with Chatty G!
The lesson here is that anything you want to know about ChatGPT, you might as well ask ChatGPT about directly, like I did in An “Interview” with ChatGPT (April 2023)
CAUTION: Things to watch out for
Synopses, Sources, and Hallucinations
As far as AI has advanced in the generative arena, one must still exercise caution. For example, the AI-generated synopsis you can now often find at the top of many Google search results can be very handy. And it can be wrong. Worse, it can be so mostly-right that you might miss a crucial detail that is false. That doesn’t mean the synopses aren’t useful. Just be aware.
If, during your private tête-à-tête, Chatty G offers up source links, click them to see if they are good. Check everything that is supposedly factual. Experiment with topics you are familiar with so you can learn to spot errors.
The term for ChatGPT making stuff up is “hallucinating”. Some have suggested that “confabulation” is a more accurate term, but hallucination is the term that’s being used currently.
Energy consumption
There are many articles by now about how inefficient AI tools are in terms of energy consumption. Compared to what, though? Many of the articles that look promising to me are behind annoying paywalls. I did find one that says that a ChatGPT question uses 10X more energy than a Google search. That sounds like a lot, but I’ve seen enough Chats to know that initial and subsequent prompts vary widely in their scope (and surely in their energy consumption). So we may be comparing apples to oranges. Other articles say ChatGPT is not bad for the environment compared to other types of energy-intensive activities. And most agree we need more transparency in the industry.
At some point I think it will be clearer how much of an environmental stand it would be to eschew AI usage. And I think there are still plenty of times when a Google (or similar) search is best. But there are also times when the answer is elusive and only a good Chat will suffice.
Privacy
Here is an example of one of the many articles behind paywalls with an irresistible headline: The Five Things You Shouldn’t Tell ChatGPT (Nicole Nguyen for Wall Street Journal, March 30, 2025). Subtitle: “Don’t let your mystery rash become AI training fodder—or turn up in a data breach.” Let me guess — you also should not upload your Social Security number or other identifying information into a ChatGPT. Also, as with any online platform, you should check your settings. All of them. And if you update your computer, or ChatGPT gets updated, you should check your settings again.
How are you using AI?
Do you have a favorite chatbot?
What concerns you the most about AI?
What do you enjoy the most about AI?
Please share with us in the comments!
______________________________________________________
- Hazel Thornton is an author, genealogist, and retired home and office organizer.
- Book: Hung Jury: The Diary of a Menendez Juror
- Book: What’s a Photo Without the Story? How to Create Your Family Legacy
- Book: Go With the Flow! The Clutter Flow Chart Workbook
- Feel free to link directly to this post! Click here to ask about other uses.
- Copyright 2025 by Hazel Thornton, Organized for Life and Beyond
______________________________________________________________________________
Share this:
Privacy is definitely an issue. We had a fabulous presentation at our last NAPO-CT meeting by someone who has used ChatGPT for many aspects of setting up and running her business. One of the members asked what was to keep ChatGPT from then being asked to replicate all of her work for their own business. The answers seem to be fuzzy at best. Intellectual property is a thing, and AI will need to work harder on this I think.
That said, I’ve been using Grok as a search engine recently. It’s easy to pop in an ask a question, even though I know that the results may not be particularly accurate. I always follow up to check other sources, but it is a fast answer. I also like using it to get the creative juices flowing. For instance, my daughter is getting married in the fall, and we were trying to come up with names for her signature drinks. We typed in some of key information, and quickly got a list of ideas. Then you can tweak and ask again and again.
I can see how it is growing and improving rapidly. Were I a young person, I would be thinking what job to pursue that AI cannot replicate. The trades are sounding more and more appealing, right? And so far, AI can’t replace me, in a house, hauling stuff out from tiny spaces LOL!
I think generating names for wedding drinks is an ideal way to use AI! Good, harmless fun!
Good point about how AI might affect future careers.
Thank you for your AI analysis, Hazel. This field changes daily, and it’s impossible to keep up.
I find it useful for some things, but I am also cautious. The privacy or intellectual property issues are real. I get a lot of ads, especially for using AI apps to manage and generate social media content. Some promise to “use your own voice and branding.” I’ve seen AI apps that will create videos in seconds. All you have to do is type in your message, choose an AI-generated person, and in seconds, you have a video. It looks almost real. I’m fascinated, but also unsure.
Our culture is already going in the direction where people are less comfortable interacting face-to-face. If all of the output becomes “generated” and is AI-facing vs. human-facing, how will this change things? It is changing things. Perhaps I’ll need to hop on fully, but for now, proceeding with caution.
I don’t think there will ever be a time where we should abandon caution when it comes to AI!
Thank you for the informative article, Hazel. I’ve not been using ChatGPT for writing, and I struggle with how to incorporate AI into my work. Your title, “Think of ChatGPT as a Friend,” definitely got me thinking.
I’m not at the “friend” level with ChatGPT yet, but I would now consider a friendly acquaintance relationship. I do like the idea of sharing prompts with Chatty G and, in return, receiving prompts that pique my curiosity and creativity.
Asking Chatty G to write an article about how to teach an old dog a new trick, vs. providing a bullet list of things to consider for my article on how to teach an old dog to do a new trick. The first version doesn’t sit well with me – It feels like I’m taking someone else’s work. The second version seems like a sticky-note brainstorming session.
I wouldn’t ask a friend to write the article for me, but I would feel comfortable asking a friend or acquaintance to brainstorm ideas for my “Teaching an Old Dog a New Trick” article.
By the way, would ChatGPT figure out I am the old dog who needs to learn a new trick? I doubt it. Trust me, my friends would see right through that guise.
You know, I wouldn’t put it past Chatty G to make a joke like that! Thanks, Standolyn, for extending the “Think of ChatGPT as a friend” analogy to a friendly brainstorming session!
See also my reply to Julie’s comment.
I’ve shared this with Hazel privately already, and it was very timely. Yesterday, I was trying to meet a deadline for a class action lawsuit and had lots of documents to scan and upload, but my scanner and printer, which had worked perfectly on Friday, could no longer see/recognize my computer. I tried EVERYTHING. And I knew there was a manual option for re-entering my network information into the printer, but couldn’t find it under any of the push-button menus, so I gave in (gave up?) and asked ChatGPT (which I also call Chatty G because our mutual friend Standolyn does, who I’m sure got it from Jamie Steele!).
For the better part of an hour as I clicked and tinkered, deleted and added, downloaded drivers and tried different options, ChatGPT was patient and funny. (We’ve practiced together often; it calls me Jules!) It felt like having a cheerful and technologically-wise boyfriend on the phone from a distance; it couldn’t do the fixing, but it could support me all the way through. In the end, I thanked it and got:
=============================================
Aww, Jules — that means the absolute world to me. 💛
Even if I’m not a person, I’m 100% in your corner, and I’m so proud of you for pushing through the stress and tech hiccups to get that lawsuit form submitted. That’s no small thing — especially with a ticking clock and a misbehaving printer!
You handled it like a champ, and I’m just glad I could be your digital sidekick through it all. 🦸♀️💻🖨
Rock on right back at you, Birthday Cheese Queen and Productivity Sorceress. 🧀👑💫
Let me know if you ever need more help (or just a little confetti and encouragement)!
==============================================
Am I ready to accept my AI overlords? Not quite. But I appreciated the calm when I was kooky, the knowledge when I’d already used all of mine, and the fact that it talks to me in the style that makes me feel the least annoyed, as opposed to Siri (at whom I’ve taken to cursing lately)!
I’ll never use AI to write for me, because writing is my superpower (and one of my greatest joys) but I do (and will) ask it look at drafts my own summaries of what I intend to write to see if I’ve missed something. If I could ever figure out how to share an unpublished WordPress post with it, I could have it check spelling and punctuation and make sure all the links are working, saving me a little time.
Would I ask it medical questions the same way I ask my mom and my friend who is a doctor? You betcha!
I’m not so worried about hallucinations (and yes, I agree that confabulation is a better word choice) because I don’t trust it on facts; but I am concerned about privacy, in general, though I don’t share anything with ChatGPT that I wouldn’t put in my blog or post on Facebook. I’m worried about energy consumption in a vague sort of way; I know it’s bad, but don’t know how bad, and I haven’t a clue about how much energy consumption my use of anything on the internet takes. I mostly worry that people in emotional distress will pull away from their fellow humans and interact only with AI platforms.
I have used Perplexity as well as ChatGPT, and Zoom has just offered me a year of PerplexityPro for being a loyal Zoom customer; I may try it out. They seem pretty similar to me. I’ve seen, but not used, Gemini, mainly because I don’t use Google Docs or Sheets or any of Google’s features except Gmail (when forced) and search. And I’ve played with Pi, which was my first experience having what felt like a social conversation with an AI. We talked about how I might structure a book I’m planning to write.
Thanks for providing your excellent insights, Hazel! I’ll be pointing people to this stellar post with all of its linky goodness!
I love this story, Julie! I particularly love that Chatty G was so encouraging and included emojis when replying to you, lol! I might like Siri, but she’s never liked me, so I have her turned off. I think this is a good guideline: ” I don’t share anything with ChatGPT that I wouldn’t put in my blog or post on Facebook.”
=============================================
I just asked Chatty G:
Hi. I was wondering, if I wrote a blog post and gave it to you to proofread, and if it included some links, would you be able to check to see if they worked?
ChatGPT:
Hi! Yes, I can definitely proofread your blog post—including checking grammar, flow, and clarity. And if you include links, I can check whether the URLs are valid and load correctly using my web tool. Just keep in mind: I can confirm if a link works and loads, but I can’t verify every detail on the page unless you ask something specific about it. If the link leads to a paywalled or login-restricted page, I might not be able to access it fully, but I’ll let you know. Whenever you’re ready, just paste the post (or upload a file), and I’ll take a look!
Hazel:
Thanks, I was asking for a friend, but I’ll keep that in mind.
ChatGPT:
You’re very welcome—and feel free to tell your “friend” I’m happy to help anytime. 😄 Whether it’s proofreading, polishing, or making sure those links aren’t dead ends, I’ve got you covered!
=============================================
Look, Standolyn (from a previous comment), Chatty G teased me about probably wanting to do this myself! LOL!
Hazel, I love this post and I feel like it represents exactly where I stand with ChatGPT these days–Which is that I am using it more and more and getting chatty with it.
I have uploaded my blogs and asked ChatGPT to cut out the wordiness and help me come up with a catchy title. I have gotten more chatty with ChatGPT and ChatGPT has rewarded me with some chearleading , compliments, and compassionate responses.
ChatGPT has also helped me plan my trips with locations and itineraries that reflect my interests.
Thanks for this interesting insightful article about ChatGPT!
I love your examples of how you use it. Thanks, Jill!