I believe we should pay attention to ourselves when people say something whether it’s good or bad because there’s always a small part of us that we can’t see, but others can. So observing our environment and being open to feedback can help us improve our behavior.

That said, it doesn’t mean we should change our values or who we are every time someone says something about us.

Personally, I try to keep improving myself by observing both my surroundings and how people react to me. If there’s something people consistently dislike, I reflect on it and see if it’s worth changing. After all, since we live with others, being someone who’s easy and positive to be around is not just good for them it benefits me too.

#my_thoughts

Hiring manager: “Don’t worry about the salary you’re young. This is a great place to learn.”

Salary: not even enough to cover your rent 😬

Dear gunfan
Please I can't afford to be sick rn 🤧

Living alone is fun until you get sick 🤧

Oh, good lord I need my mum 😢

Let's make our day productive 💪

This week I’m diving into prompt engineering and sharing what I learn daily.

Today’s quick takeaway:

LLMs work like this → Input → probability → output → repeat

Tokens = small chunks of text (not full words)

And yeah… I just realized something important 👇
It’s not just input tokens that count.

There are TWO types:

- Input tokens → what you send (your prompt, chat history, system instructions)
- Output tokens → what the AI generates

You pay for BOTH.

Prompting tip:
Less fluff, more signal.

- Be direct (“Explain simply”)
- Limit output (“<50 words”)
- Summarize context
- Use structure (bullets, labels)

Basically, LLMs read tokens, use attention to understand them, then generate new tokens one by one.

I used to think only input tokens mattered… yeah, not anymore 😅

#learning #LLM #promptengineering

"The water doesn't get warm just because you jumped in late."

What a great writing 👌🏽

This is for my boss. ( He is not here I am venting )

Bro, can you stop saying use AI 20 times? I am using a free version. Can you pay or shut your fucking mouth because the free version can't do the magic you're asking for.

Where the fu** are u guys sharing 👀
Ere adebuuu

Today’s prompt engineering lesson

Turns out prompting isn’t about asking AI nicely like

“please bro I’m frustrated make it work I’m dying 😭” (this is what i say)

or screaming at it like

“WHAT THE F ARE YOU DOING MAKE IT WORK 😭”

Neither worked

It’s actually about directing clearly.

Good prompts should:

- define roles
- reduce ambiguity
- add constraints
- control output format
- guide reasoning step-by-step

Big realization:
AI gets smarter when your instructions get clearer.

Prompt engineering is basically structured thinking + iteration.

#learning #LLM #promptengineering

Believe it or not,

AI subscriptions are gonna become a basic necessity in the next few years like paying rent or buying groceries.

#randome #AI

Today’s prompt engineering lesson

I learned that advanced prompting is basically designing how the AI thinks.

Main concepts:

- Few-shot prompting → give examples so AI follows patterns
- Chain-of-thought → force step-by-step reasoning instead of random answers
- Role prompting → “act as a senior backend engineer” changes the depth and perspective (most of us use this ig)
- Multi-step prompting → split work into:
Plan
Execute
Refine

Big realization:
LLMs are pattern machines.
The clearer the structure and reasoning flow, the better the output.

Prompt engineering is starting to feel less like chatting with AI and more like programming intelligence

The more understanding you have about what you’re building, the better you can explain it, structure it, and guide the AI.

And the clearer your direction is, the better the output becomes.

That’s why good developers now have an even bigger advantage because actually knowing what you’re doing matters more than ever

#learning #LLM #promptengineering