Today I started my 90-day journey into Artificial Intelligence.
No hype, no shortcuts — just learning step by step and sharing what I understand.
What I Learned Today
I began with Large Language Models (LLMs).
Before today, I genuinely thought AI understands language like we do.
But the reality is very different.
LLMs don’t understand language.They predict the next word.
How It Actually Works
At its core, an LLM is trained on massive amounts of text data.
Given an input like:
"AI is changing the..."
The model predicts the most probable next word, such as:
"world"
It does this by learning patterns and probabilities from data — not by understanding meaning.
Every sentence generated is just a sequence of these predictions.
What Surprised Me
Even without true understanding, LLMs can:
Write essays
Answer questions
Generate code
This made me realize:
Intelligence (in this case) can emerge from simple rules applied at scale.
My Takeaway
What feels like “thinking” is actually:
Pattern recognition
Probability
Massive training data
And somehow, that’s enough to create powerful AI systems.
What’s Next?
Tomorrow I’ll explore:
What tokens are
How text gets converted into numbers
Day 1 complete. 89 days to go.
If you’re also learning AI, let’s connect and grow together.
Top comments (0)