You've reached the final level. Today we dive into the code, secure the digital world, and learn how to use AI responsibly to change the world.
Computers don't just follow instructions; they learn from experience.
import sklearn
# Step 1: Gather Data
data = [[150, "Red"], [170, "Orange"], [130, "Yellow"]]
# Step 2: Select a Model
model = sklearn.linear_model.DecisionTree()
# Step 3: Train the Machine!
model.fit(data, labels)
# Step 4: Make Predictions
print(model.predict([[160, "Red"]]))
Learning with a teacher. We give the AI the answers during training.
The AI finds patterns in messy data all by itself!
Learning through rewards. Like training a puppy with treats!
Natural Language Processing is how AI understands human emotions and intent.
Breaking sentences into small chunks called "Tokens". AI doesn't see words; it sees numbers and probabilities.
Determining if a sentence is Happy, Sad, or Angry. Companies use this to see if customers are happy with their products.
Using AI to solve the world's biggest problem: Climate Change.
Drones use AI to see exactly which plant needs water, saving 90% of water compared to traditional farming.
Smart grids use AI to predict when a city will need electricity, reducing waste and carbon emissions.
AI cameras in forests can detect poachers or monitor endangered species 24/7 without disturbing them.
In Class 10, we don't just build models; we evaluate them using math.
Percentage of correct guesses. (Correct / Total).
Of all predicted positives, how many were actually correct? Useful for spam detection.
Of all actual positives, how many did the AI catch? Critical for medical diagnosis.
Generative AI doesn't just copy; it "blends" concepts in a mathematical space called Latent Space. Slide to mix two different faces!
Moving the slider is like moving through the Latent Space of a GAN model.
IB Criterion D: Evaluating. What happens if our AI fails?
Your AI detected cancer with 99% accuracy but missed 1 person. How do you explain this to the family? Reflect on the ethical responsibility of an AI Master.
How does your brain's structure compare to the architecture of ChatGPT?
Your brain is a massive, messy web of 100 Trillion connections. It learns by physically changing the strength of synapses over years of experience.
Modern AI uses Attention Mechanisms. It mathematically "attends" to the most important words in a sentence to understand meaning instantly.
Learn how Generative Adversarial Networks create hyper-realistic art.
An AI that tries to create "fake" images to fool the Critic.
An AI that acts like a judge, trying to spot the fake image.
This "War" between two AIs is how tools like Midjourney create stunning visuals!
Watch how an AI learns! Lower the Loss (error) to make it smarter.
Accuracy: 0%
Current Loss: 2.50
Adjusting weights... minimize the loss to converge.
AI is powerful, but it can be used for the wrong reasons.
Using AI to swap faces in videos or mimic voices. This makes it hard to know what is real online.
If we train AI on "unfair" data, the AI will also be unfair. This is why Diversity in data is so important!
Test your coding logic! What will be the output of this Python code?
x = 5
if x > 3:
print("AI")
else:
print("Robot")
With great power comes great responsibility. Secure your AI!
AI needs data, but it must never steal your personal secrets. Always use "Anonymized Data".
Hackers can trick AI by showing it "noisy" images. We must build robust models to stay safe.
AI decisions affect real lives. How would you choose?
A self-driving car is about to hit a sudden obstacle. It must choose between:
Always use AI to assist your learning, not to cheat! If you use AI to write an essay, mention it in the credits. Honesty is the first step to becoming a great engineer.
How does AI create art? It uses a Generative Adversarial Network.
The "Artist" AI that tries to create a realistic image from scratch.
The "Critic" AI that tries to guess if the image is real or fake. They compete until the Generator is perfect!
Move the slider to see how AI interpolates between two features!
Feature A
Feature B
Adjusting the Latent Space vector...
Using AI to save our planet.
Drones use AI to spray water only where plants are thirsty. Saves 60% of water!
AI predicts when people need electricity most and balances the grid to avoid blackouts.
Robotic boats use Computer Vision to find and collect plastic in the ocean.