What I Learned from Google’s Generative AI Course (A Beginner-Friendly Breakdown)
Hey, I’m Shaila 👋 A CS student navigating the ups and downs of coding, learning, and building meaningful projects. If you’re learning step by step, figuring things out as you go, you’ll feel right at home here.
I recently completed a Google course on Generative AI, and it helped me clearly understand how different AI concepts connect—from basic machine learning to modern generative models.
This article is a simple breakdown of what I learned, written from a learner’s perspective.
AI Is a Discipline, Not Just a Tool
One of the first things I learned is that Artificial Intelligence (AI) is a broad discipline. It’s not a single technology or model, but a field focused on building systems that can perform tasks requiring human-like intelligence—such as reasoning, learning, and decision-making.
Within AI, there are several subfields, and one of the most important is Machine Learning.
Machine Learning: Learning from Data
Machine Learning (ML) is a subfield of AI where systems learn patterns from data instead of being explicitly programmed. ML is mainly divided into two categories:
1. Supervised Learning
Uses labeled data
Each input comes with a correct output
The model:
Takes input data (x)
Predicts an output
Compares it with the expected output
Calculates an error if the prediction is wrong
Improves itself using that error
Example: Predicting house prices when the dataset already includes correct prices.
2. Unsupervised Learning
Uses unlabeled data
The model looks at raw data and tries to find patterns on its own
Common tasks include clustering and grouping similar data points
Example: Giving the model customer data and letting it discover natural groups without telling it what those groups mean.
In simple terms, unsupervised learning is about discovering structure in data.
Deep Learning: A More Powerful Form of Machine Learning
Deep Learning is a type of machine learning that uses artificial neural networks to process complex patterns in large datasets.
Neural networks are inspired by the human brain and consist of:
Input layer
Hidden layers
Output layer
Each layer processes information and passes it forward, allowing the model to learn increasingly complex representations of data.
Because of this layered structure, deep learning models perform extremely well in tasks like image recognition, speech processing, and natural language understanding.Labeled, Unlabeled, and Semi-Supervised Learning
Neural networks can work with:
Labeled data
Unlabeled data
Or a combination of both
When a model uses:
A small amount of labeled data to learn basic concepts
And a large amount of unlabeled data to generalize and generate new patterns
This approach is called Semi-Supervised Learning.
It’s especially useful when labeled data is expensive or difficult to obtain.
Generative AI: Creating New Data
Generative AI is a subset of Deep Learning.
Unlike traditional models that only analyze or classify data, generative models can create new data that resembles what they were trained on—such as text, images, or audio.
Generative vs Discriminative Models
The course also introduced two important types of models:
Discriminative Models
Focus on distinguishing between different kinds of data
Learn the boundary between classes
Example: Spam vs Not Spam classifiers
Generative Models
Learn how the data itself is structured
Can generate entirely new data instances
Examples: Text generation, image generation, music generation
Large Language Models (LLMs)
Large Language Models are also a subset of deep learning and generative AI.
They are trained on massive amounts of text data and function as pattern-learning systems. Based on the input you provide, they:
Identify patterns
Predict the most likely next words
Generate coherent and context-aware responses
They don’t “understand” language like humans—but they are very effective at learning patterns from data.
How can we distinguish whether data is generative or not?
Final Thoughts
This course helped me see AI not as a buzzword, but as a structured ecosystem:
AI → Machine Learning → Deep Learning → Generative AI
Each layer builds on the previous one.
This foundational clarity makes advanced AI topics feel far less intimidating—and much more exciting to explore. Learning the basics first creates a strong base for understanding everything that follows.
