• AIpeeps
  • Posts
  • Learning Loop ♾️ 07-2025

Learning Loop ♾️ 07-2025

Your bi-weekly dose of learning AI, without the hype.

Hello peeps,

Hello everyone, and a very warm welcome to the 7th edition of AIpeeps!

What we’re covering today:

  • Experimenting with Claude for PPTs – testing if it really beats ChatGPT

  • Prompt Amplifier – making prompts 100% better

  • Google’s Nano Banana – quick edits, sharp results

  • Jailbreaking ChatGPT – what it is, why people try, and why I don’t recommend it

💡3 Curious things I learnt

1. Experimenting with Claude for PPTs

Lately, I’ve been hearing a lot that Claude performs way better than ChatGPT for PPTs and Excel.
So, I gave it a try.

Here’s the prompt I used 👇

The way it processed things and displayed every aspect of formatting was mind-blowing.

Now, let’s see the results:

  • With ChatGPT, PPT formatting feels quite basic. It can’t really handle much, and I’m sure your experience has been the same.

  • With Claude, the output was 100% better. The thought process was also much clearer.

Yes, I did get a limit error because I was using the superior model and it got exhausted.

👉 You can check the result yourself in the link below.

One thing I noticed: it didn’t pull all the information from my blog.
Still, I’d say — try Claude for your next PPT. With ChatGPT, I’m losing hope for formatting.

2. Prompt Amplifier

I call it “Prompt Amplifier” because it makes your prompt 100% better and helps you start with the right input.

There’s a saying: if you start off well, you always have a good chance to succeed.
The same applies to prompts.

So, I tested it with the same PPT prompt from earlier, but this time amplified. Here’s how it looks:

How it works:

  1. It asks for your basic prompt.

  2. Then, it asks you to add specific details.

Finally, it creates a refined prompt for better results.

Yes, it took a lot more time to arrive at the final prompt.

👉 I’ll share the amplified Claude results and final prompt in my next newsletter.

3. Google – Nano Banana

I already use ChatGPT Plus, but I still explore other tools. Why?

Because these niche tools sometimes deliver mind-blowing results for their specific use cases.

Tried Google’s Nano Banana recently → quick edits, sharp output.

You’ll even see the pic it generated in my email signature today.

🕵🏻Decoding the Jargon

Jailbreaking ChatGPT

Think of ChatGPT like a student in class with rules to follow:

  • Teacher says: “Don’t share test answers, don’t draw on the walls, don’t say mean things.”

  • These are the safety guardrails for ChatGPT.

Now, jailbreaking means trying to bypass or trick it into ignoring those rules.
People do this with clever prompts to make ChatGPT do things it normally wouldn’t — like:

  • Unsafe instructions

  • Revealing hidden data

  • Breaking usage policies

⚠️ My advice → Don’t try it.


Yes, there’s a whole Reddit group dedicated to jailbreaking ChatGPT, but it’s not worth the risk.

Try it Out this week:

✔️ Try making a PPT with Claude.

✔️ Use a Prompt Amplifier to sharpen your prompts.

✔️ Avoid jailbreaking ChatGPT — it’s risky and unnecessary.

If you learned something today and liked the newsletter — take a screenshot and post it on LinkedIn with your honest feedback. Tag @Gaurav Jain (or drop the link in a DM) — I’ll pay for your 1-month ChatGPT Plus subscription (only for existing subscriber till 16.09.2025).

Share the love, flex what you learned, and let’s grow this convo. 🚀

Please feel free to share any questions or topics you're particularly curious about.

Hope you learned something new today!

Till next time,

Reply

or to participate.