Domain 1 of 4

Fundamentals of Gen AI

The vocabulary, the three ways a model learns, the machine learning lifecycle, how a foundation model is chosen for a use case, what data quality means in practice, the five layers of the landscape, and Google's own models.

Concepts

~30%

Of the exam

Practice questions

Concepts in this domain

01Supervised, unsupervised and reinforcement learningThe three ways a model learns, what each one needs from your data, and how to tell which problem you are actually holding.02The machine learning lifecycleThe stages between a business question and a model in production, and why the work does not stop at deployment.03Foundation models and how they workTokens, embeddings and transformers explained for somebody who has to make decisions about a model rather than build one.04Choosing and configuring a foundation modelThe criteria that actually separate one model from another, and what temperature, top p and output length do to the answers you get back.05Data for generative AIWhat makes data usable by a model, the difference between structured and unstructured and between labelled and unlabelled, and why accessibility beats volume.06The five layers of the gen AI landscapeInfrastructure, models, platforms, agents and applications, and why knowing which layer you are buying at decides what you are responsible for.07Google's foundation modelsGemini, Gemma, Imagen and Veo, what each is for, and how to pick between an open model you host and a managed one you call.

Try a question from this domain

A retailer wants a customer service assistant that answers from its own returns policy, checks an order in its warehouse system, and issues a refund where the policy allows it. Which layer of the gen AI landscape does this describe?

AModels, because a foundation model produces every response.
BAgents, because the system plans steps, calls tools and acts rather than only answering.
CInfrastructure, because it requires compute to serve requests.
DApplications, because customers interact with it directly.

18 questions on this domain.

One per page, with a worked explanation.

Start the set