Practice Exam: Salesforce AI Associate #1

Question 1

What is the main difference between predictive AI and generative AI?

Accepted Answer

Predictive AI analyzes historical data to make predictions; generative AI creates new content like text and images

Answer

Predictive AI creates new content; generative AI classifies existing data

Answer

Generative AI exclusively uses neural networks; predictive AI uses only basic statistics

Answer

Both do exactly the same thing but with different commercial names

Question 2

A company discovers that its AI hiring model systematically rejects candidates from certain geographic areas. What type of AI risk does this represent?

Accepted Answer

Bias

Answer

Hallucination

Answer

Toxicity

Answer

Poor Grounding

Question 3

Which are Salesforce Trusted AI principles? (Select 2)

Accepted Answer

Answer

Accountability

Answer

Maximization of economic benefits

Answer

Transparency

Answer

Processing speed

Question 4

What is a 'hallucination' in the context of generative AI?

Accepted Answer

When AI generates information that sounds plausible but is factually incorrect or fabricated

Answer

A hardware error on the server running the model

Answer

When the model takes too long to respond

Answer

A training technique to improve model accuracy

Question 5

What function does Einstein Trust Layer serve in Salesforce?

Accepted Answer

Protect customer data and privacy when interacting with AI models, masking PII and preventing data retention by external LLMs

Answer

Completely replace the need for human review

Answer

Store all model training data

Answer

Manage Salesforce user licenses

Question 6

What is 'grounding' in the context of Salesforce generative AI?

Accepted Answer

Anchoring AI model responses in specific, contextual organizational data to improve relevance and accuracy

Answer

Connecting server hardware to physical ground

Answer

Optimizing model performance by reducing its size

Answer

Deleting all previous training data from the model

Question 7

Which are key components of data quality for AI? (Select 3)

Accepted Answer

Answer

Accuracy: data reflects reality

Answer

Completeness: records don't have critical empty fields

Answer

Consistency: data follows uniform formats and standards

Answer

Antiquity: the older the data, the better

Question 8

An AI chatbot starts generating responses containing insults and offensive language. What AI risk is manifesting?

Accepted Answer

Toxicity

Answer

Bias

Answer

Hallucination

Answer

Data Leakage

Question 9

What is Supervised Learning?

Accepted Answer

A type of ML where the model learns from labeled data with known correct answers

Answer

A model that works without any training data

Answer

An approach where a human manually supervises each model prediction in production

Answer

A system that only works with unstructured data

Question 10

What is PII (Personally Identifiable Information) in the AI context?

Accepted Answer

Data that can directly or indirectly identify an individual, such as name, email, social security number, etc.

Answer

A type of AI algorithm for image processing

Answer

An integration protocol between AI systems

Answer

The preferred data format for training Machine Learning models