Muhammad Ikram Let's talk

Ever wonder how ChatGPT ACTUALLY works?

Here's the plot twist: It doesn't "think" — it predicts.
Just created a breakdown explaining the real mechanics behind ChatGPT:
-> What tokens really are and why they matter
-> Training vs. Inference — the difference between learning and performing
-> Why hallucinations happen (spoiler: confidence doesn't equal accuracy)
-> The truth: It's not magic, it's sophisticated pattern prediction

Think of ChatGPT as autocomplete on steroids, trained on the entire internet, playing the world's most elaborate game of "What word comes next?"

It doesn't remember. It doesn't understand. It predicts based on patterns.
And somehow, that's enough to revolutionize how we work and create.
Check out the full slides to understand what's really happening under the hood of the AI everyone's talking about.

Read the slides as text

Slide 1

Ever Wonder How ChatGPT
Ever Wonder How ChatGPT
Actually Thinks? 🤔
Actually Thinks? 🤔
🤖💭
Spoiler: It doesn't. But let me explain why that's both hilarious and mind-blowing!
Not the generic "AI is smart" stuff...
Let's dive into the REAL mechanics!

Slide 2

Here's The Plot Twist 😱
Here's The Plot Twist 😱
🎲🎯
Imagine autocomplete on steroids, trained on the entire internet, and trying REALLY hard to sound
smart.
ChatGPT doesn't "think"
It's basically playing the world's most elaborate game of "What word
comes next?"

Slide 3

🧩 What Are Tokens?
🧩 What Are Tokens?
Think of tokens as ChatGPT's vocabulary puzzle pieces
Why it matters: ChatGPT doesn't see words, it sees these chunky little tokens. It's like trying to
read a book where every word is broken into syllables!
📚➡️🧩🧩🧩
"Hello" = 1 token (easy peasy)
→
"ChatGPT" = 2 tokens (Chat + GPT)
→
"Antidisestablishmentarianism" = Like 7 tokens (it's having a moment)
→

Slide 4

📖 Training Data: The Ultimate Binge Read
📖 Training Data: The Ultimate Binge Read
Then someone asks you a question, and you desperately try to remember what you read...
"But you don't actually REMEMBER — you just predict what would
SOUND right based on patterns."
Imagine reading:
Every Wikipedia article ever written
→
Millions of books (the good, the bad, the Reddit threads)
→
Countless websites (yes, probably that embarrassing blog post you wrote in 2012)
→

Slide 5

🎓 Training vs. Inference
🎓 Training vs. Inference
TRAINING (The Gym Phase) 💪
Months of reading everything
→
Learning patterns: "After 'Once upon a' usually comes 'time'"
→
Costs millions of dollars in compute
→
Happens ONCE (or occasionally with updates)
→
INFERENCE (Game Time) 🎮
That's when YOU talk to ChatGPT
→
It uses what it learned to predict responses
→
No new learning happens here!
→
It's like taking an open-book test... without the book
→

Slide 6

💣 Here's The Kicker
💣 Here's The Kicker
"During inference (when you chat), ChatGPT can't learn or remember new
facts. It's frozen in time!"
⏰❄️
It's like talking to someone with movie-level amnesia, but they're REALLY good at pretending they
remember!
So when you tell ChatGPT your name?
It's not "remembering" — it's just keeping your conversation in its short-term memory (context
window). Once that chat is over... *poof* 💨

Slide 7

🌈 Why Hallucinations Happen
🌈 Why Hallucinations Happen
The million-dollar question!
"It's like that friend who's NEVER been to Paris but describes it in detail
based on movies they've seen."
🎭✨
Remember: ChatGPT predicts words, it doesn't "know" facts
Pattern over truth: If something "sounds right" based on patterns, it'll say it
→
Confidence without knowledge: It can sound SUPER sure about completely made-up stuff
→
The autocomplete trap: Sometimes the statistically likely next word isn't the TRUE next word
→

Slide 8

🎯 The Bottom Line
🎯 The Bottom Line
"It doesn't know. It predicts.
And somehow... that's enough to blow our minds! 🤯"
ChatGPT is NOT:
Thinking 🧠
→
Remembering 💭
→
Understanding 🤓
→
Knowing "the truth" 📚
→
ChatGPT IS:
An incredibly sophisticated pattern
predictor 🎯
→
Really, REALLY good at sounding
convincing 🎭
→
Completing your sentences at scale 📝
→
Next time ChatGPT surprises you, remember: it's not magic, it's math. Really impressive, mind-bogglingly complex math! 🧮
✨

More insights

Keep reading

All posts
Carousel

AI & Tech Explained 1 min read

AI Pilots Impress. Only Production Moves Revenue.

A common trap: the AI demo wins the room, then stalls. Gartner predicted at least 30% of GenAI projects would be dropped after proof of concept. The growth lesson: budget the data, owner, integration and run cost before the demo.

AI & Tech Explained 1 min read

Stripe, Adyen or Checkout.com?

Stripe, Adyen or Checkout.com? The right payment provider depends on where your business is going—not just the brand name. There’s no universal winner.…