Chatbots Chatbot 3 — Questions and Answers
Question 1: What is a generative AI chatbot primarily based on?
- Large language models that produce novel text (Correct answer)
- A fixed list of canned answers
- A spreadsheet of replies
- A keyword-matching table only
Correct answer: Large language models that produce novel text
Generative chatbots use large language models to create original responses.
Question 2: What is 'training data' for a chatbot?
- The examples used to teach the model how to respond (Correct answer)
- The bot's color scheme
- The hosting bandwidth
- The user's IP address
Correct answer: The examples used to teach the model how to respond
Training data consists of example conversations and text that shape the model's behavior.
Question 3: What is a 'hallucination' in the context of AI chatbots?
- A confidently stated but false or fabricated answer (Correct answer)
- A slow loading screen
- A successful login
- A user complaint form
Correct answer: A confidently stated but false or fabricated answer
A hallucination is when the model generates incorrect information presented as fact.
Question 4: What does a 'prompt' refer to when interacting with an AI chatbot?
- The input or instruction given to the model (Correct answer)
- The bot's loading animation
- A type of firewall
- The output log file
Correct answer: The input or instruction given to the model
A prompt is the text or query a user provides to elicit a response.
Question 5: Which approach improves chatbot accuracy by retrieving relevant documents before answering?
- Retrieval-Augmented Generation (RAG) (Correct answer)
- Random sampling
- Cookie tracking
- Load balancing
Correct answer: Retrieval-Augmented Generation (RAG)
RAG fetches relevant external information to ground the model's response.
Question 6: What is a key advantage of chatbots for customer service?
- 24/7 availability and instant responses (Correct answer)
- They never need updates
- They eliminate all human staff
- They guarantee zero errors
Correct answer: 24/7 availability and instant responses
Chatbots can respond instantly at any hour without queues.
Question 7: What is 'sentiment analysis' used for in chatbots?
- Detecting the emotional tone of a user's message (Correct answer)
- Measuring server temperature
- Counting message length
- Compressing files
Correct answer: Detecting the emotional tone of a user's message
Sentiment analysis identifies whether a user feels positive, negative, or neutral.
Chatbots Chatbot 3
What is a generative AI chatbot primarily based on?