- AIpeeps
- Posts
- Learning Loop ♾️ 04-2025
Learning Loop ♾️ 04-2025
Your bi-weekly dose of learning AI, without the hype.

Hello peeps,
Hello everyone, and a very warm welcome to the 4th edition of AIpeeps!
This week I am bit late sending the newsletter, I was improving the readability of your learning letter.
What we’re covering today:
Temporary Chat – to keep things off the record
Deleting Memory – deleting what ChatGPT remembers about you
Study & Learn – GPT-5 that teaches instead of just answering.
Vector- space world of words
💡3 Curious things I learnt
1. Temporary Chat in ChatGPT
ChatGPT now has a Temporary Chat option — basically incognito mode for your conversations.
It’s tucked away so subtly you might miss it, but it’s exactly what many of us have been waiting for.
Why?
Keeps certain chats out of your history.
Perfect for when you don’t want anyone tracking what you were searching.
Avoids awkward situations, if someone happens to see your screen.
How to find it- Right corner-side - you will see chat sign as shown below:

Temporary Chat
2. Deleting ChatGPT Memory
Yes — you can delete memory.
How memory is created:
During each chat, ChatGPT notes your qualities, preferences, or current challenges. Over time, it builds a profile to give more tailored responses.
To check:
Ask “What do you know about me?” — you’ll get a direct summary.
To delete:
Go to Settings → Memory → Manage Memories.
Select what you want to remove and press the bin icon.
This ensures ChatGPT won’t use that context again.

Manage Memory
💡 Note: ChatGPT does not have unlimited memory.
3. Use “Study and Learn” in ChatGPT
Think of this as ChatGPT switching from answering to teaching.
Instead of just giving you the result, it:
Explains step-by-step
Uses examples & analogies
Shares why the concept matters
Can include visual data or plots
Ends by asking you questions to confirm your understanding
The goal? You don’t just get an answer — you actually learn the concept.

Study and Learn
🕵🏻Decoding the Jargon
Vector
In physics, a vector means something with both size (magnitude) and direction — like an arrow showing how far and which way.
But in the world of LLMs, “vector” is about words living in a kind of invisible space.
Each word (or chunk of text) gets turned into a vector with thousands of dimensions — way more than we can picture.
In that space, words that mean similar things end up close together, and words that mean very different things are far apart.
That’s how the model can figure out meaning, connections, and context.
Try it Out this week:
✔️ Try Temporary Chat for sensitive or one-off searches
✔️ Use Memory deletion if you don’t want certain info stored
✔️ Experiment with Study & Learn in GPT-5, plus test features like Think Deeply and Vibe coding
I'm truly excited to learn alongside you all.
Please feel free to share any questions or topics you're particularly curious about.
Hope you learned something new today!
Till next time,
Curiously yours,
Gaurav Jain
Reply