Before you start: what this course is
This course prepares you for the Google Cloud Generative AI Leader certification exam. Three facts first:
- This course is independent of Google. It is not produced, endorsed or reviewed by Google.
- It does not guarantee a pass. It teaches what the published exam guide covers; the result depends on your study.
- The checks in this course are scenario-based practice questions written for learning. They are not real exam questions.
No technical background or coding is needed. The exam is aimed at any job role and tests business-level knowledge.
Task 1: Memorise the exam format
Learn these figures now so you can plan your time.
- Questions: 50-60, all multiple choice
- Length: 90 minutes
- Prerequisites: none
- Delivery: online-proctored or onsite-proctored
- Sections: four, each with an approximate weighting
| Section | Name | Approx. weighting |
|---|---|---|
| 1 | Fundamentals of gen AI | ~30% |
| 2 | Google Cloud's gen AI offerings | ~35% |
| 3 | Techniques to improve gen AI model output | ~20% |
| 4 | Business strategies for a successful gen AI solution | ~15% |
Do this:
- Note that Sections 1 and 2 together make up roughly 65% of the exam. Plan most of your study time there.
- Work out your pace: 90 minutes for up to 60 questions is about 90 seconds per question.
- Check Google Cloud's official certification page before you book. Fees, languages and policies can change.
Task 2: Learn the four nested terms
The exam expects you to tell four terms apart. Each one sits inside the one before it, like a set of nesting boxes.
- Artificial intelligence (AI) is the largest box. It is the broad field of building systems that perform tasks normally requiring human intelligence, such as reasoning, understanding language or recognising images.
- Machine learning (ML) sits inside AI. Instead of following rules written by a person, an ML system learns patterns from data and uses them to make predictions or decisions on new data.
- Deep learning sits inside ML. It uses neural networks with many layers to learn complex patterns from large volumes of data such as text, images and audio.
- Generative AI (gen AI) sits inside deep learning. It produces new content (text, images, audio, video, code) based on patterns learned from existing data.
Read the nesting in one direction only:
- All gen AI is deep learning, all deep learning is ML, and all ML is AI.
- The reverse is false. Not all AI is ML (a hand-written rule system is AI but learns nothing from data). Not all ML is gen AI.
Do this: write the four terms from broadest to narrowest from memory: AI → ML → deep learning → gen AI.
Task 3: Separate generative tasks from predictive tasks
Many scenario questions turn on one distinction: does the business want new content, or a label or number?
- Predictive ML outputs a label, category, score or number about existing data. Typical verbs: classify, detect, forecast, predict, score.
- Generative AI outputs new content that did not exist before. Typical verbs: write, draft, summarise, create, design, compose.
| Business need | Output | Task type |
|---|---|---|
| Flag card transactions as fraud or not fraud | A label | Predictive (classification) |
| Estimate next quarter's sales | A number | Predictive (forecasting) |
| Decide which customers are likely to cancel | A score | Predictive |
| Draft a reply to a customer complaint | New text | Generative |
| Produce product images for a campaign | New images | Generative |
| Summarise a 40-page contract | New text | Generative |
Do this for every scenario:
- Find the output the business wants.
- Ask: is it a label or number, or is it new content?
- Label or number means predictive ML. New content means gen AI.
Task 4: Place natural language processing (NLP)
Natural language processing (NLP) is the area of AI that enables computers to understand, interpret and produce human language, written or spoken. In business terms: NLP is what lets software work with the words people actually use, in emails, reviews, chats and calls.
How NLP relates to the other terms:
- NLP and AI: NLP is a field within AI. It is defined by its subject (human language), not by a single technique.
- NLP and ML: modern NLP relies on machine learning, and mostly on deep learning, to learn language patterns from data.
- NLP and large language models (LLMs): an LLM is a deep learning model trained on very large amounts of text. LLMs are the technology behind most current NLP. A later level covers them in detail.
- NLP and gen AI: the two overlap but are not the same.
- NLP that is not generative: sorting reviews into positive or negative (sentiment analysis), or routing emails by topic. The output is a label.
- NLP that is generative: drafting an email, summarising a report, answering a question in a chat.
- Gen AI that is not NLP: generating an image or a video.
Do this: when a scenario involves language, apply the Task 3 test as well. Language plus a label is predictive NLP. Language plus new text is generative.
What the exam asks about this
Expect scenario-based multiple-choice questions at business level. For this level's material, be ready to:
- Pick the correct definition of AI, ML, deep learning, gen AI or NLP from four similar-sounding options.
- Identify which term is the broader or narrower one.
- Read a short business scenario and decide whether it describes a generative or a predictive task.
- Recognise that a language task can be either predictive or generative.
Key takeaways
- The exam has 50-60 multiple-choice questions in 90 minutes, across four sections weighted ~30%, ~35%, ~20% and ~15%.
- This course is independent of Google and promises no pass.
- Nesting order, broadest to narrowest: AI → ML → deep learning → gen AI.
- Predictive ML outputs a label or number; gen AI outputs new content.
- NLP is the field of AI that deals with human language. It uses ML and deep learning, is powered today largely by LLMs, and overlaps with gen AI without being identical to it.