Your AI Wrapped: What ChatGPT Knows About You (And What You Can Do About It)

Recently I started seeing social media posts of people asking ChatGPT (or their AI of choice) to give them a “Spotify Wrapped” version for their AI usage.

So of course, I had to try it because I love my Spotify Wrapped and I love my AI.

I dropped this prompt into ChatGPT:

Prompt 1 – “AI Wrapped”
give me my yearly summary as if you’re spotify wrapped! tell me the most unhinged things i said this year and give me a forecast for the year ahead!

What came back was…a lot.

I didn’t even remember asking half the things it surfaced. But as I read, I could feel my body going, “Oh right, I did ask for that. And that. And…oh dear lord, that too.”

It was a vivid reminder that:

• I go to AI way more often than I consciously realize.

• I use it as a thought partner on a wide range of topics.

• These platforms are quietly building a picture of who I am and how I think through my prompts.

And no, I am not as bad as this lol

It felt like it remembered everything. (Technically, it’s a mix of saved “memories” plus chat history(more on that below), but the effect is the same: it knows a lot.)

This turned out to be a really important exercise: a slightly fun, slightly uncomfortable mirror on my actual AI habits and not the habits I think I have.

And it’s making me want to dig much deeper into data, memory, and what happens to all the information we keep feeding these tools.

Prompt 2: Ask ChatGPT to Analyze Your Usage

After the “AI Wrapped” experiment, I wanted something more structured, less entertainment, more insight.

So I asked ChatGPT to act as an analyst of my usage, using only what it could see from my history and metadata. Here’s the exact prompt I eventually developed after several iterations to arriving at one that I liked:

Prompt 2 – “Usage Analyst”
`I’d like you to act as an analyst of my ChatGPT usage.

Using ONLY what you can see or infer from:

• my past conversations in this account,

• any long-term memory or usage metadata you have access to about me (such as frequency of use, typical conversation depth, and common topic types),

create a clear, human-readable profile of how I tend to use ChatGPT.

Please structure your response in these sections:

1. Usage Patterns

   • How often I seem to use ChatGPT (if you can infer it).

   • Typical conversation depth (do I ask quick questions or have multi-step, iterative dialogues?).

   • Anything you can see about message length or how much context I usually provide.

2. Typical Tasks & Themes

   • The main types of things I use ChatGPT for (e.g., writing help, lesson design, coding, planning, personal reflection, decision support, etc.).

   • Any recurring themes, domains, or topics you notice.

3. My Working Style With ChatGPT

   • How I seem to “work” with the model (e.g., do I iterate, refine, ask for multiple versions, request structure, etc.?).

   • Any patterns in how I give instructions (very detailed vs. very brief, etc.).

4. Strengths in How I Use ChatGPT

   • What I appear to be doing particularly well in my usage (e.g., giving context, iterating, leveraging structure).

5. Opportunities to Use ChatGPT Even Better

   • 4–6 concrete, practical suggestions tailored to my patterns.

   • Focus on things like: better prompt structures, reusable templates, ways to save time, or ways to get higher-quality thinking out of the model.

Important guidelines:

• If you do not have access to a particular kind of data (for example, exact usage frequency), say that clearly instead of guessing.

• Base everything on evidence from my conversations, not generic advice.

• Keep the tone direct and constructive, as if you’re an executive coach giving me an honest usage review.`

The result was basically an “executive debrief” of how I use AI: what I do well, what I overuse it for, and very specific ways I could be using it more intentionally.

I actually prompted it next to create a one pager that I then turned into an infographic using Google Nano Banana

Questions to Sit With After You Run These

If you run either or both prompts, don’t just skim the output. Use it as a reflective tool:

• What surprised you the most about how you actually use AI?

• Where are you outsourcing thinking that you might want to pull back and do more slowly or analog?

• Where are you not using AI that could lower your cognitive load or give you back time?

• If a student or colleague saw your “AI Wrapped,” what story would it tell about your priorities and values?

• What boundaries (topics, tasks, decisions) do you want to set for yourself going forward?

If you’re feeling brave, you might even share your AI Wrapped or usage insights with a trusted colleague or team and talk about what it reveals.

Okay, But What About All This Data?

Once you see how much the system can reconstruct about you, it’s natural to ask:

• What exactly is it keeping?

• Can I see it?

• Can I delete it?

With ChatGPT today, there are three important buckets to understand:

1. Saved Memories – details the system stores about you (preferences, projects, bio, etc.) and reuses in future conversations. OpenAI Help Center

2. Chat History – the transcripts of your conversations, which can also be referenced to inform new replies. OpenAI Help Center

3. Data Used to Improve the Model – whether your chats are used to train models (“Improve the model for everyone”) is controlled separately in your Data Controls. OpenAI Help Center

You have quite a bit of control over all three if you know where to look.

How to See and Search What ChatGPT Remembers About You

Here’s a practical walkthrough you can share with staff, students, or families.

1. Ask Directly: “What do you remember about me?”

You can simply type:

What do you remember about me?

ChatGPT will list the high-level memories it’s actively using (e.g., your role, interests, recurring projects). This is a good “quick x-ray,” but it’s not the whole picture, but there’s also a dedicated Memory area in Settings. OpenAI Help Center

2. Open the Memory Settings and Search

On the web (signed in):

1. Click your profile picture.

2. Go to Settings → Personalization → Manage memories. OpenAI Help Center

   

3. There you can:

   • See a list of what’s been saved about you.

   • Search your memories and sort by newest or oldest.

   • Delete individual memories (⋯ → Delete) or delete all memories at once. OpenAI Help Center

   • Turn Memory on or off entirely.

Important nuance: turning Memory off stops new memories from being stored, but doesn’t automatically erase existing ones as you still need to delete them in Manage memories. OpenAI Help Center

3. Fully Deleting Something: Two Places to Clean

If there’s a particular fact you want gone (e.g., where you live, a project name, a personal detail), you need to:

1. Delete the relevant memory in Settings → Personalization → Manage memories. OpenAI Help Center

2. Delete any chat(s) where you originally shared that information.

Only doing one of these can leave traces behind; doing both is the cleanest option. OpenAI Help Center

How to Delete or Archive Chats

To clean up specific conversations:

1. In the sidebar, hover over the chat title.

2. Click the three dots (⋯).

3. Choose Delete to remove it or Archive to tuck it away. OpenAI Help Center

Deleted chats disappear from your history and are then scheduled for permanent deletion from OpenAI’s systems (with some exceptions for security/legal reasons). You can’t get them back once they’re gone. OpenAI Help Center

Archiving keeps the chat out of your main view but doesn’t delete it.

Data Controls You Should Check Periodically

Under Settings → Data Controls, you can: OpenAI Help Center

• Turn off “Improve the model for everyone.”
  Your chats still work and still appear in your history, but they’re excluded from model training.

• Export your data.
  Get a copy of your ChatGPT data for your records or to audit what’s there.

• Delete your account.
  The nuclear option, but it’s there if you ever need it.

For extra-sensitive conversations, you can also use Temporary Chats, which:

• Don’t appear in your history.

• Don’t create memories.

• Aren’t used to train models and are deleted after 30 days. OpenAI Help Center

A Simple “AI Hygiene” Checklist

Here’s a lightweight routine you (or your staff/students) could adopt:

• Once a quarter:

  • Run the “AI Wrapped” and/or “Usage Analyst” prompts.

  • Reflect: How is my AI usage shifting? Is that what I want?

• Once a month:

  • Visit Settings → Personalization → Manage memories.

  • Search and delete any memories you no longer want stored.

  • Decide whether Memory being on or off matches your current comfort level.

• Any time you’re doing something sensitive:

  • Consider using a Temporary Chat instead of your main history.

  • Double-check Data Controls if you don’t want your chats used for training.

• Once or twice a year:

  • Export your data from Data Controls if you want a personal record.

  • Do a bigger cleanup of old chats you no longer need.

If you end up trying either of these prompts and get a response that makes you laugh, cringe, or rethink how you’re using AI, I’d love to hear about it.

We talk a lot about “AI literacy,” but a big part of it is this: being honest about our own patterns, and learning how to manage the data trail we’re leaving behind.