Ai6 Flashcards
What does AI stand for?
Artificial Intelligence
True or False: AI is only used for robots.
False
What does ML stand for in AI?
Machine Learning
What is the main difference between AI and ML?
Simple example:
AI: A robot that can clean your house and chat with you.
ML: The part of that robot that learns which rooms get dirty fastest based on past cleaning data.
What does NLP stand for?
Natural Language Processing
Fill in the blank: NLP helps computers understand ________.
human language
What does API stand for?
Application Programming Interface
Which best describes an API?
“A set of rules that lets software programs communicate with each other”
True or False: LLM stands for Long Language Machine.
False
What’s is an LLM
An LLM (Large Language Model) is a type of AI trained on huge amounts of text so it can understand and generate human language.
How to remember it:
Think of an LLM like a giant reservoir filled with words. When you ask it something, it draws from that reservoir — like turning on a tap — and gives you a stream of language instead of water.
What does SDK stand for?
Software Development Kit
Fill in the blank: An SDK gives developers the ________ they need to build software.
An SDK (Software Development Kit) gives developers the tools, libraries, documentation, and code samples they need to build software for a specific platform or framework.
What does a vector database store?
Numbers that represent meaning, used for fast AI search.
What is a common use of embeddings in AI?
To compare the meaning or similarity between words, documents, or images by converting them into numerical vectors that capture semantic relationships — like plotting ideas on a map, where similar meanings are closer together.
Fill in the blank: A chatbot like ChatGPT is powered by a ______.
Large Language Model (LLM)
What is the goal of a classification problem in AI?
To sort data into categories, like spam vs not spam.
Detailed explanation:- To automatically assign input data to one of several predefined categories or classes based on patterns learned from labelled training data. For example, identifying whether an email is spam or not spam, or classifying images of animals into categories like cats, dogs, or birds.
Regression problems are used when you want to predict a ________.
number
True or False: An AI that predicts house prices is likely using regression.
True
What is the main role of training data?
To help the AI learn by showing examples.
What does RAG stand for?
Retrieval-Augmented Generation
Fill in the blank: RAG helps combine LLMs with your own ________.
data
What is the purpose of prompt engineering?
To write better inputs that guide an AI to give useful answers.
True or False: Prompt engineering only works with voice assistants.
False
What is supervised learning?
Learning from data that includes correct answers.
What is unsupervised learning?
Learning from data without labels, finding patterns.
What is an example of classification?
Classification is when an AI model assigns inputs into predefined categories. A common example is classifying emails as “spam” or “not spam”. Other examples include identifying whether an image contains a cat or a dog, predicting if a customer is likely to churn or stay, or determining the sentiment of a product review as positive, negative, or neutral.
What is overfitting in AI?
When a model learns training data too well and fails on new data.
True or False: Overfitting means the model performs well on unseen data.
False
Fill in the blank: A neural network is made up of many connected ________.
layers
What does ‘inference’ mean in AI?
Using what the model learned to make predictions.
Which of these is an example of inference in the context of AI?
A. A model is trained on a dataset of cat and dog images.
B. A model predicts an email is spam based on its content.
C. A dataset is cleaned and labelled.
D. A neural network is being designed.
B. A model predicts an email is spam based on its content.
⸻
Inference in AI refers to the process of using a trained model to make predictions or decisions based on new data. Option B is the only one where the model is applying what it has learned—i.e., performing inference.
What is an embedding?
A number-based representation of something like a word or image.
Fill in the blank: Embeddings help AI understand ________ rather than just exact words.
meaning
True or False: A vector database can help an AI find similar documents.
True
What does it mean to label data?
To mark examples with the correct answers for training.
Which of these is unsupervised learning?
A) Sentiment analysis
B) Clustering customers by behaviour
C) Image recognition
B) Clustering customers by behaviour
Explanation:
Unsupervised learning finds patterns in unlabelled data. Clustering groups similar customers without pre-defined labels.
What’s the difference between training and inference?
Training is learning from data, inference is using what it learned.
Fill in the blank: Classification sorts things into categories, while regression predicts a ________.
number
What is a real-world example of classification?
A bank deciding if a loan application is low or high risk.
What’s a real-world example of regression?
Estimating the price of a used car based on age and mileage.
True or False: A model trained with supervised learning doesn’t need labelled data.
False
What’s the main job of a neural network?
To process information and find patterns in data.
True or False: An LLM like ChatGPT needs no training data.
False
What is fine-tuning?
Adjusting an AI model using your own data to make it more accurate.
What’s a real-world example of RAG?
Using ChatGPT to answer questions based on company documents.
Why do AI models use so much data?
Because more data helps them learn patterns more accurately.
What is a variable in Python?
A name used to store a value, like a label for information.
Fill in the blank: In Python, name = ‘Andrew’ is storing the word ‘Andrew’ in a ________ called name.
variable
What is a Python function?
A block of code that does something. You can use it again and again.
True or False: Functions can take inputs and return outputs.
True
What does a for loop do in Python?
It repeats an action for each item in a list or sequence.
What is a list in Python?
An ordered group of items, written in square brackets.
What is a dictionary in Python?
A group of key-value pairs, like a contact list.
True or False: Python is the most popular language for AI.
True
What is a Jupyter notebook used for?
Writing, testing, and sharing code — often used for data science and AI.
What is Hugging Face?
A website where you can find and use AI models.
What is AutoML?
Tools that help people build machine learning models without much coding.
Fill in the blank: The ‘OpenAI API’ lets you send ________ to an AI model and get answers.
prompts
What is tokenisation in AI?
Breaking text into small pieces (tokens) so an AI can understand it.
True or False: AI models charge based on the number of tokens used.
True
What does ‘context window’ mean for an LLM?
The maximum amount of info it can remember at once.
What does ‘inference cost’ mean?
How much it costs to run an AI model on new data.
Final Q: What’s the main benefit of understanding these fundamentals?
So you can use AI more effectively in real life, even without being a programmer.
