Chapter 5

Question 1

what is AI

Answer 1

Artificial Intelligence (AI) is the field of computer
science dedicated to creating machines that can perform tasks requiring human intelligence

Question 2

key areas of AI

Answer 2

Machine Learning, NLP, Computer Vision, Robotics, Expert Systems

Question 3

what are the types of AI

Answer 3

*Narrow AI
*General AI
*Super AI

Question 4

What is Narrow AI

Answer 4

Designed for a single task
(e.g., Siri, facial recognition).

Question 5

what is General AI

Answer 5

Hypothetical system with human-level intelligence.

Question 6

what is Super AI

Answer 6

Theoretical AI surpassing human intelligence.

Question 7

what is generative AI

Answer 7

A subfield of AI focused on creating original content and new data.

Question 8

what is the core function of generative AI

Answer 8

Learns dataset patterns to generate new, similar data

Question 9

examples of generative AI

Answer 9

• Text: Articles, summaries, code, poetry.
• Images: Art, photos, animations.
• Audio: Music, voice cloning, sound effects.
• Video: Deepfakes, synthetic footage.

Question 10

history of AI
* 1950s-1980s

Answer 10

: Rule-based simple text/music generation.

Question 11

history of AI
* 1980s-2010s

Answer 11

: Neural networks enabled complex pattern
generation.

Question 12

history of AI
* 2014

Answer 12

: GANs revolutionized image generation.

Question 13

history of AI
* 2017-Present

Answer 13

: Transformers enabled GPT and advanced
LLMs.

Question 14

Key Generative Techniques

Answer 14

1. Generative Adversarial Networks (GANs)
2. Diffusion Models
3. Transformers

Question 15

what is Generative Adversarial Networks (GANs):

Answer 15

This technique uses a two-part system to create content.
The Generator creates the content, while the Discriminator
judges it, and they “compete” to improve the quality of the
output.

Question 16

can create realistic, but fake, images of human faces that don’t belong to any real person.

Answer 16

GANs - Generative Adversarial Networks

Question 17

what is Diffusion Models

Answer 17

These models work by taking an image and adding noise to it. They then learn to reverse this process, which allows them to create a high-quality image from random noise.

Question 18

You can use this technique to generate a unique digital painting from a text prompt like, “a serene landscape with a river and mountains at sunset.”

Answer 18

diffusion model

Question 19

what is Transformers

Answer 19

The backbone of modern Large Language Models (LLMs). They use an “attention” mechanism to weigh the importance of different words in a sentence, giving them a deep understanding of context.

Question 20

this technique like GPT can write a summary of a long article, generate computer code, or compose a poem based on your request.

Answer 20

transformer model

Question 21

The backbone of modern LLMs

Answer 21

Transformers

Question 22

what is LLM

Answer 22

Large Language Models

Question 23

what is the AI life cycle

Answer 23

Data Gathering = Collecting clean datasets
Model Training = Computationally intensive learning phase
Inference/Generation = Producing outputs from prompts
Fine-tuning & Deployment = Task optimization and release

Question 24

Everyday Uses of Generative AI (At School)

Answer 24

• Study Aids: Summaries and flashcards.
• Research Assistance: Topics, sources, organization.
• Tutoring: Personalized practice.
• Language Learning: Dialogues, grammar explanations.
• Creative Assignments: Scripts and project ideas

Question 25

Ethical and Social Considerations

Answer 25

1. Misinformation & Deepfakes. 2. Bias. 3. Copyright & Plagiarism. 4. Job Displacement.

Question 26

what is Misinformation & Deepfakes:

Answer 26

AI has the potential to create convincing, but fake, content. This includes deepfakes, which are synthetic videos or images where a person's face or voice is digitally altered, often for deceptive purposes. This can lead to the spread of misinformation.

Question 27

what is bias

Answer 27

AI models can amplify existing biases found in their training data.

Question 28

if a model is trained on a dataset of job applications that are predominantly male, it might learn to favor male candidates, leading to prejudiced outputs.

Answer 28

bias

Question 29

what is Copyright & Plagiarism

Answer 29

There are legal and ethical questions about the ownership of content created by AI. It's also unclear how to address situations where AI-generated content closely resembles existing copyrighted works.

Question 30

what is Job Displacement

Answer 30

As AI becomes more advanced, there is a concern that it could automate tasks previously done by humans, leading to changes in the job market and potential job displacement. This could affect creative fields like graphic design and writing, as well as more routine tasks.

Question 31

Limitations of Generative AI

Answer 31

1. Hallucinations 2. Lack of True Understanding 3. Dependency on Training Data 4. Copyright and Plagiarism Concerns 5. Environmental Impact

Question 32

Generative models can produce plausible-sounding but factually incorrect information. This is often referred to as

Answer 32

hallucinations

Question 33

what is hallucinations

Answer 33

make up information because their goal is to create a coherent response, not a factually accurate one.

Question 34

These models are excellent at recognizing patterns in data, but they don't possess genuine consciousness or comprehension. They are essentially complex pattern matching machines that can't reason like humans.

Answer 34

Lack of True Understanding:

Question 35

The quality and scope of a generative AI's output are entirely dependent on the quality and diversity of the data it was trained on.

Answer 35

Dependency on Training Data

Question 36

dependency on training data can lead to a few issues:

Answer 36

*Bias: If the training data contains biases (e.g., gender, racial), the model will learn and replicate them. *Lack of Real-time Information: Most models have a knowledge cutoff and can't access or discuss events that have happened since their last training.

Question 37

Because the AI learns from existing content, there are legal and ethical questions about the ownership of content it creates and whether it infringes on existing copyrighted material

Answer 37

Copyright and Plagiarism Concerns

Question 38

Training massive generative AI models requires a huge amount of computing power and energy, which raises concerns about their carbon footprint.

Answer 38

Environmental Impact

Question 39

The Future of Generative AI

Answer 39

1. Multimodal AI 2. Real-time Integration 3. Regulation

Question 40

what is multimodal AI

Answer 40

Future models will seamlessly integrate and understand multiple forms of data, such as text, images, audio, and video, at the same time. This will allow for more dynamic and creative outputs.

Question 41

You could give a model a text prompt, a piece of music, and an image, and it would generate a short, animated video that combines all of these elements.

Answer 41

multimodal AI

Question 42

what is Real-time Integration

Answer 42

Generative AI will become more deeply embedded in everyday applications and devices, providing instant, personalized assistance without noticeable delay.

Question 43

A future version of a virtual assistant could instantly summarize a live lecture you are attending or automatically generate meeting notes and action items in real time

Answer 43

Real-time Integration

Question 44

what is Regulation

Answer 44

As the technology becomes more powerful and widespread, there will be an increased focus on developing policies and ethical frameworks to ensure its safe, responsible, and equitable use.

Question 45

Governments or international bodies may create new laws that require AI models to be transparent about their training data to help mitigate bias and copyright issues.

Answer 45

regulation