Intro

Question 1

What is Artificial Intelligence

Answer 1

AI is the theory and development of computer systems able to perform tasks normally requiring human intelligence. AI is a discipline, like how Physics is a discipline of science.

Question 2

What is Machine Learning

Answer 2

A sub-field of AI, It is a program or system that trains a model from input data. The trained model can make useful predictions from new (never-before-seen) data drawn from the same one used to train the model.

This means that Machine Learning gives the computer the ability to learn without explicit programming.

Question 3

Supervised ML models

Answer 3

Supervised models have labelled data a.k.a data with a tag that identifies that data.

Question 4

Unsupervised ML models

Answer 4

Unsupervised models have unlabelled data which is harder to work with vs labelled data

Question 5

Supervised learning implies..

Answer 5

that the data is already labeled. In supervised learning we are learning from past examples to predict future values.

Question 6

Unsupervised learning implies..

Answer 6

that the data is not labelled. Unsupervised problems are all about looking at the raw data, and seeing if it naturally falls into groups.

Question 7

What is the unsupervised learning system flow?

Answer 7

Input data -> Model -> Generated Example

Question 8

What is the supervised learning system flow?

Answer 8

Input data -> Model -> Predict Output -> Compare to expected output from training data -> Error -> update model

Question 9

What is Deep Learning?

Answer 9

A sub-set of machine learning. Deep learning uses Artificial Neural Networks - allowing them to process more complex patterns than traditional machine learning.

Question 10

What is Generative AI?

Answer 10

Generative AI is a subset of Deep Learning, & a type of AI that creates new content based on what it has learned from existing content. the process of learning from existing content is called training and results in the creation of a statistical model. When given a prompt, GenAI used this statistical model to predict what an expected response might be - and this generates new content.

Question 11

What are Large Language Models?

Answer 11

LLMs are also a subset of Deep Learning

Question 12

What 2 types can Deep Learning Models be divided into?

Answer 12

Discriminative & Generative

Question 13

What is Discriminative AI?

Answer 13

Used to classify or predict, it’s typically trained on a dataset of labeled data, and learns the relationship between the features of the data points and the labels. E.g classifies a dog from a photo

Question 14

What is Generative AI as a model?

Answer 14

Generates new data that is similar to data it was trained on, Understands distribution of data and jow likely a given example is, and predicts next words in a sequence. E.g generates a picture of a dog

Question 15

What outputs are not Gen AI outputs?

Answer 15

Numbers, Discrete, Class, & Probability

Question 16

What outputs are Gen AI outputs?

Answer 16

Natural Language, Images, Audio

Question 17

Generative Language Models

Answer 17

Generative language models learn about patterns in language through training data. Then, given some text, they predict what comes next.

Question 18

Generative image models

Answer 18

Generative image models produce new images using techniques like diffusion. Then, given a prompt or related imagery, they transform random noise into images or generate images from prompts.

Question 19

What are Hallucinations?

Answer 19

Hallucinations are words or phrases that are generated by the model that are often nonsensical or grammatically incorrect.

Question 20

What causes Hallucinations?

Answer 20

The model is:
- not trained on enough data
- trained on noisy or dirty data
- not given enough context
- not given enough constraints

Question 21

What is a Prompt & Prompt Design?

Answer 21

A prompt is a short piece of text used to control the output of the LLM by inputting it into the LLM to “prompt” the model. Prompt Design just means that the quality of the input determines the quality of the output.

Question 22

What different model types are there?

Answer 22

text-to-text

Question 23

What is the text-to-text model type?

Answer 23

Text-to-text models take a natural language input and produce text output. These models are trained to learn the mapping between a pair of texts.

Uses include:
- text generation
- text classificiation
- text summarization
- text translation
- text (re)Search
- text extraction
- text clustering
- Content editing / rewriting

Question 24

What is the text-to-image model type?

Answer 24

Text-to-image models are relatively new and are trained on a large set of images, each captioned with a short text description. Diffusion is one method used to achieve this.

Uses include:
- Image generation
- Image editing

Question 25

What are the text-to-video & text-to-3D model types?

Answer 25

Text-to-video models aim to generate a video representation from text input. The input text can be anything from a single sentence to a full script, and the output is a video that corresponds to the input text. Similarly Text-to-3D models generate three-dimensional objects that correspond to a user’s text description (for use in games or other 3D worlds).

Question 26

What is the text-to-task model type?

Answer 26

Text-to-task models are trained to perform a specific task or action based on text input. This task can be a wide range of actions such as answering a question, performing a search, making a prediction, or taking some sort of action. For example, a text-to-task model could be trained to navigate web UI or makes changes to a doc through the GUI.

Question 27

What is a foundational model?

Answer 27

A large AI model pre-trained on a vast quantity of data, designed to adapted or fine-tuned to a wide range of downstream tasks such as:

• question answering
• sentiment analysis
• information extraction
• image captioning
• object recognition
• instruction following