The Curious Companion: Ep. 23 – Using ChatGPT’s Personalization Settings to Get (Way) Better Responses

Curious Reader!

Welcome to this week’s Curious Companion newsletter.

What you came for is below, and you can CLICK HERE to listen to the episode if you decide you prefer earbuds to eyeballs.

Happy reading!

In this episode, I share how a small, easily ignored setting completely changed ChatGPT’s responses…in a good way. I discuss the personalization settings, including base tone, custom instructions, memory, and advanced settings, and go into detail about how “programming” ChatGPT how to respond immediately made for a WAY better user experience. If you’ve ever thought “this thing is fine but kind of annoying,” this episode is for you.

Holiday + Life Updates

In an effort to inject some humanness and timeliness into these episodes, let me start by saying Merry Christmas to those of you who celebrate, happy holiday season to those of you who celebrate something else, and happy Thursday to those of you who don’t celebrate anything.

Christmas isn’t a big deal here. Thanksgiving is the big holiday, so Lex and I will be chilling. It’s also my brother’s birthday, so a very happy birthday to him. He doesn’t read these, but happy birthday nonetheless.

Vacation, Vibe Coding, + the Missed Feature

Lex and I just got back from vacation in Playa del Carmen and it was doooooope.

During that time, I watched some AI and vibe coding videos from one of the self‑paced online courses I’m taking, and one of the lessons brought up something I have not done: Using the custom instructions field in the personalization section of ChatGPT. Inset face-palm.

In episode 9, I outlined a bunch of the personalization features, but I 100% skipped over the custom instructions field. Mainly because I don’t like free‑type questions, and for whatever reason I told myself that what I put in there would be ignored. I was wrong.

An Intro to Personalization Settings

To access the personalization section of ChatGPT, click on your name at the bottom of the main menu on the left side of the screen. From there, click personalization, and a new window will open.

The first option is a dropdown menu that lets you select the base style and tone.

There are eight options: default, professional, friendly, candid, quirky, efficient, nerdy, and cynical. For those of you wondering, I use candid.

Custom Instructions (The Actual Game Changer)

The next section of personalization settings is what I really want to highlight and what I’ve been skipping: custom instructions.

What has been objectively and significantly impactful for me is using this area to tell ChatGPT how I want it to respond. This is a free‑type space, and you’re limited to 1500 characters.

To generate this text, I had ChatGPT interview me.

The prompt I used was:

I’d like you to interview me one question at a time so you can generate custom instructions that I will paste into that personalization section in ChatGPT.

If you have no idea what you want to focus on in terms of instructions and guidelines, I’d suggest starting there.

What I found as it started interviewing me was that I did in fact know what I wanted. I wanted instructions related to brevity, avoiding and minimizing hallucinations, minimizing sycophancy, and avoiding that initial fluff nonsense it generates when providing answers.

The follow‑up prompt I gave it was:

These questions are not helpful. Mainly I want you to interview me so you can create instructions related to brevity, avoiding and minimizing hallucinations, sycophancy, and just giving me responses without the initial fluff like, “That’s a great question, Maestro. Here’s the answer without any fluff.”

It is my hope that by now you have realized that you can speak to Chat like a person. You can have typos. You can use regular language. Yes, it’s doing math to give you an answer, but how you can communicate with it, is, IMO, the real magic.

Refining the Instructions

After that reframe, the questions were much better, and it actually gave me both the questions and bulleted list options for the answers. I didn’t prompt that response style, but you can if it’s not giving you what you want.

After asking the questions, it generated paste‑ready custom instructions, but they were too long, which is how I learned about the 1500‑character limit. I asked it to regenerate said instructions, and to do so without losing any of the requirements we established.

From there, I pasted the final version into the custom instructions field, and honestly, I’m loving the change.

The responses are streamlined and to the point. It flags when it’s making inferences. It’s not agreeable just to be agreeable (which is fucking dope to me), and it will actually disagree with me.

How you set it up will be up to you, but I can confirm that what goes in the custom instructions box does in fact matter and does change ChatGPT’s behavior.

The “About You” Section

Back to the personalization settings.

The next section is “About You”. In this section there are are free‑type fields for nickname, occupation, and more about you, such as interests, values, or preferences to keep in mind.

I actually have these blank. You’d think I’d learn based on my recent revelation about how impactful the custom instructions box was, but ChatGPT already knows these things about me based on three years of usage. I do however think that answering these fields would be really helpful right out of the box for new users.

Memory Settings

The next section, which I spoke about in episode 9, is Memory.

This section has a manage button where you can view a list of all the things ChatGPT remembers about you. In this area you can delete anything you don’t want Chat to remember. I think it’s really cool to go through this section, and I strongly suggest you take the time to see what Chat has stored about you in its memory.

To add to memory, you can simply use the prompt “add to memory” whenever you’re using Chat and want it to remember something about you. It will add it to the saved memory bank.

The rest of the Memory section is toggle fields, and there are explanations for each setting. For those wondering, I have reference saved memories on, referenced browser memories off, and reference chat history on.

Record Mode + Advanced Settings

There are two more sections in the Personalization setting: Record Mode and Advanced.

Record mode is a single on or off toggle. I have it toggled on.

The advanced section is an accordion, and when you click it, a handful of options appear, each with an explanation. I have all of them toggled on.

That rounds out the entirety of the personalization settings. Of note, you can change the voice for voice mode, but that lives in general settings, not personalization.

Worth it?

It is well worth spending time in the personalization section, especially custom instructions and managing memory.

It has been my experience over the past week that it absolutely makes for a better user experience, and I think it’s only going to continue to improve as OpenAI focuses more on memory and personalization in order to maintain market share and attempt to differentiate themselves from other LLMs.

How I Used ChatGPT This Week

Each episode, I include a section where I briefly discuss how I used ChatGPT that week.

This week, I used ChatGPT to help me redesign my workout split. I’m pretty sure I’ve mentioned this use case before, but I leaned on it again and had Chat do a lot more this time.

Last time, Chat was just a sounding board. This time, I had ChatGPT create the entire program, and then we edited it together.

Worth noting, and tying back to this episode, at one point I prompted ChatGPT to remember that I’m an advanced lifter and have a doctorate in physical therapy, which was helpful in how it communicated with me and the exercises it chose.

Da Wrap-up

Hopefully this episode has nudged you to spend some time in ChatGPT’s personalization settings, especially custom instructions and memory. It can absolutely be a game-changer.

As always, endlessly appreciative for you and your curiosity.

Catch you next Thursday.

Maestro out.

AI Disclaimer: In the spirit of transparency (if only we could get that from these tech companies), this email was generated with a very solid alley-oop from ChatGPT. I write super detailed outlines for every podcast episode (proof here), and then use ChatGPT to turn those into succinct, readable recaps that I lightly edit to produce these Curious Companions. Could I “write” it all by hand? Sure. Do I want to? Absolutely not. So instead, I let the robot do the work, so I can focus on the stuff that I actually enjoy doing and you get the content delivered to your digital doorstep, no Airpods required. High fives all around.

​Click here to join the ChatGPT Curious newsletter family.​

Stay curious.​