Generative AI Leader Fundamentals of Generative AI Practice Question
Which TWO are benefits of using pre-trained foundation models instead of training from scratch?
⚠ Common exam trap
Google Cloud often tests the misconception that pre-trained models eliminate the need for any further engineering (like prompt engineering) or that they are completely bias-free, when in fact they still require careful tuning and can perpetuate biases from their training data.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Lower training cost
Option B (Lower training cost) is correct because pre-trained foundation models let you skip the massive compute, data, and energy expense of pre-training from scratch, requiring only comparatively cheap fine-tuning or inference. Option E (Faster time to deployment) is correct because starting from an already-trained model means you can fine-tune or prompt it and ship a working solution in far less time than a full training pipeline. Option A is wrong because pre-trained models actually constrain architecture choices, since you inherit the provider's design rather than controlling it. Option C is wrong because prompt engineering is still typically required to steer foundation models effectively. Option D is wrong because pre-trained models can and do carry biases from their training data, so absence of bias is not guaranteed.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Complete control over model architecture
Why it's wrong here
Pre-trained foundation models ship with fixed architectures that consumers cannot redesign; fine-tuning adjusts weights, not layers. It is tempting because control over architecture matters in regulated or latency-sensitive deployments, and would be correct when training a bespoke model from scratch for a niche domain.
- ✓
Lower training cost
Why this is correct
Pre-trained foundation models remove the need for large-scale training runs from scratch, so organisations pay only for adaptation such as fine-tuning or prompting. This directly satisfies the benefit of substantially lower compute and data-labelling expenditure compared with building a model from zero.
- ✗
Eliminates the need for prompt engineering
Why it's wrong here
Prompt engineering remains necessary to steer foundation models toward desired outputs; pre-training removes data collection and training cost, not prompt design. It is tempting because these models are instruction-tuned and respond well to natural language, and would be correct only if a model required no task specification at all.
- ✗
Guaranteed absence of bias
Why it's wrong here
Pre-trained models inherit biases from their training corpora, so bias can persist or amplify after fine-tuning. It is tempting because foundation models are trained on broad, curated data, and would be correct only for models explicitly audited and debiased for a specific domain.
- ✓
Faster time to deployment
Why this is correct
Starting from a pre-trained foundation model skips months of data collection, architecture search and training, so teams reach production far sooner. This satisfies the stem's benefit of faster time to deployment, since only adaptation and evaluation remain before release.
Go deeper
Related to this question
About these practice questions
Courseiva writes every Generative AI Leader question from scratch — 1,008 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This Generative AI Leader practice question is part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the Generative AI Leader exam.