0% found this document useful (0 votes)
12 views7 pages

Generative AI - Notes

The document provides an overview of Generative AI, detailing its definition, how it works, and its various models such as GANs, VAEs, and RNNs. It discusses applications across fields like art, music, and coding, as well as the benefits, challenges, and ethical considerations associated with its use. The document emphasizes the importance of responsible use and the potential societal impacts of Generative AI.

Uploaded by

pranavarya1307
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
12 views7 pages

Generative AI - Notes

The document provides an overview of Generative AI, detailing its definition, how it works, and its various models such as GANs, VAEs, and RNNs. It discusses applications across fields like art, music, and coding, as well as the benefits, challenges, and ethical considerations associated with its use. The document emphasizes the importance of responsible use and the potential societal impacts of Generative AI.

Uploaded by

pranavarya1307
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

SILVERLINE PRESTIGE SCHOOL

SESSION - 2025-26
CLASS: 9
SUBJECT: AI
CHAPTER : GENERATIVE AI
____________________________________________________________________________

L
O
Introduction to Generative AI
●​ Generative artificial intelligence (AI) refers to the algorithms that generate new data that

O
resembles human-generated content, such as audio, code, images, text, simulations, and

H
videos.

SC
●​ This technology is trained with existing data and content, creating the potential for
applications such as natural language processing, computer vision, the metaverse, and
speech synthesis.​

E
G
How Generative AI Works
Generative AI models learn the patterns, structure, and relationships in training data and then use
TI
that learning to create new data.
ES

Two Key Learning Methods

a.​ Supervised Learning


PR

●​ Works on labelled data.


●​ Input and correct output are provided during training.
●​ Used mostly for discriminative tasks (classification, prediction).​
E
N

b.​ Unsupervised / Self-Supervised Learning


LI

●​ Works on unlabelled data.


●​ The model learns patterns without predefined answers.
ER

●​ Useful for generative modeling where the output is new content.​


LV

Conventional AI vs Generative AI
SI

Feature Generative AI Conventional AI

Goal Creates new content Analyzes, processes & classifies data

Training Use vast libraries of samples to train Employs fewer complex algorithms
neural networks and other complicated and training methods.
structures to produce new content
based on those patterns.

Output Fresh, innovative, and often Predictable, based on existing data.


unexpected.

Applications Art, music, content generation, design Banking, healthcare, recognition,


language processing

L
O
Types of Generative AI Models

O
H
1.​ GANs (Generative Adversarial Networks)
●​ GANs are neural networks that work together to generate fresh data.

SC
●​ They consist of two networks:
○​ Generator Network – creates new data.

E
○​ Discriminator Network – checks and analyzes the generated data.
●​ The discriminator provides feedback to help the generator improve.

G
●​ Both networks continue learning in a feedback loop until the generated data
TI
becomes similar to real data.
●​ Examples of GAN applications:
ES

○​ Creating portraits of non-existing people


○​ Converting images from day to night
PR

○​ Generating images from textual descriptions


○​ Creating realistic videos.​
E

2.​ VAEs (Variational Autoencoders)


N

●​ Another class of generative models is VAEs. In order to produce fresh data, VAEs
LI

learn the distribution of the data and then sample from it.
●​ Examples:
ER

○​ Generation of new images similar to a given training set


○​ Image reconstruction
LV

○​ Generating drafts for writers


○​ Generating new sounds and music composition, etc.​
SI

3.​ RNNs (Recurrent Neural Networks)


●​ RNNs are a special class of neural networks that excel at handling sequential data,
like music or text.
●​ They function by ingesting past inputs and applying that knowledge to anticipate
future inputs.
●​ Example: Generating novel text in the style of a specific author or genre,
predicting the next character or word in a sequence, etc.

4.​ Autoencoder
●​ These are neural networks that have been trained to learn a compressed
representation of data.
●​ They function by compressing data first, then decompressing that compressed
data to restore it to its original form.

L
●​ Autoencoders can be applied to denoising or picture compression applications.

O
●​ Examples: Artistic image creation, drug discovery. They generate highly realistic

O
samples.

H
Applications of Generative AI

SC
1.​ Art
●​ Used to create unique works of art.
●​ Example: The Next Rembrandt created a new painting in Rembrandt’s style.​

E
G
2.​ Music
●​ used to create new music, either by composing original pieces or by remixing
TI
existing ones.
ES

●​ Example: AIVA creates songs in different genres.​

3.​ Language
PR

●​ Used to generate new language, such as chatbots that can hold conversations with
users or natural language generation systems that can produce written content.
●​ Example: ChatGPT.​
E
N

4.​ Video Creation


LI

●​ Generates realistic animations and videos.​


ER

5.​ Game and Virtual Worlds


●​ Creates game levels, characters, and environments.​
LV

6.​ Code Generation


SI

●​ Helps write or correct programming code.​

Benefits of using Generative AI


1.​ Creativity
●​ Helps creators push boundaries and enhances creative processes.
●​ Useful in fields like art, design, and music.​
2.​ Efficiency
●​ Automates content creation.
●​ Saves time and reduces costs compared to manual work.​

3.​ Personalization
●​ Creates customized content based on user preferences and behaviour.
●​ Useful for personalized recommendations, news articles, and more.​

L
4.​ Exploration

O
●​ Helps explore new design spaces and optimize complex systems.

O
●​ Used in drug discovery, industrial improvements, etc.​

H
5.​ Accessibility

SC
●​ Makes content creation tools available to everyone.
●​ Helps people with limited skills or resources produce high-quality content.​

E
6.​ Scalability

G
●​ Generates large amounts of content quickly.
TI
●​ Suitable for businesses and organizations needing high-volume content
production.​
ES

Challenges / Limitations of Generative AI


PR

1.​ Data Bias


●​ If Generative AI is trained on biased or incomplete data, AI may produce
similarly biased or flawed results.
E

●​ This can cause inaccuracies in areas like facial recognition or language


N

processing.​
LI

2.​ Uncertainty
ER

●​ Generative AI can produce unexpected or unpredictable outputs.


●​ These unpredictable results can be both a benefit & drawback.​
LV

3.​ Computational Demands


●​ Training and generating outputs requires significant computational resources.
SI

●​ This makes the process expensive and time-consuming.

Generative AI tools
There are many generative AI tools available today that enable users to create and experiment
with generative models. Here are some popular tools:
●​ Artbreeder:
○​ Artbreeder is a web-based tool that enables users to generate new images by
combining different GAN models.
○​ Users can select and combine different GAN models to create new and unique
images.
●​ Runway ML:
○​ Runway ML is a platform for creating, training, and deploying generative models.
○​ It provides a user-friendly interface for building and training various types of
generative models, including GANs, VAEs, and image classifiers.

L
●​ ChatGPT:

O
○​ Creates text, answers questions, and helps in writing.

O
●​ Gemini:
○​ Gemini is Google’s generative AI tool that helps create text, images, answers, and

H
summaries quickly and intelligently.

SC
How to Use Generative AI Tools in Real-world Scenarios
The table shows popular Generative AI tools that can be used in various fields.

E
Category
GTools
TI
Video Muse AI, Vista AI, Vizard AI
ES

Text ChatGPT, Notion AI, Compose AI


PR

Images Midjourney, Magic Studio, Ideogram

Design Viesus, Piggy AI, Galileo AI


E
N

Coding Bugasura, CodeGPT, Replit Ghostwriter


LI

Audio FineShare, Boomy AI, Playlist AI


ER

Productivity Briefly AI, Sora AI, Leexi AI


LV

Ethical considerations of using Generative AI


While Generative AI offers many benefits, there are also several ethical considerations that
SI

should be considered when using this technology.

●​ Ownership
○​ There are ongoing questions about who owns content created by generative
AI.
○​ This issue is important in creative fields like music, literature, and art.
○​ Generative AI can produce original works that blur the boundary between
human and machine authorship.

●​ Human Agency
○​ Generative AI raises concerns about human control and decision-making.
○​ As AI becomes more advanced, it may be harder to differentiate between
human-generated and machine-generated content.

L
○​ This could reduce human autonomy and agency.

O
●​ Bias

O
○​ AI may reproduce and amplify existing biases in training data.

H
○​ Can lead to unfair decisions in hiring, loans, and justice systems.

SC
●​ Misinformation
○​ Can create fake news and deepfakes.

E
○​ Misleads people and harms public trust and democracy.

●​ Privacy
G
TI
○​ May generate or expose sensitive personal information.
ES

○​ Risk of misuse, identity theft, or data breaches.

The Potential Negative Impact on Society


PR

●​ Deep Fake or Unrealistic Content


○​ Generative AI may generate sensitive personal information.
○​ This includes details like social security numbers or medical records.
E

○​ Such information can be misused for harmful or illegal purposes.


N
LI

●​ Job Displacement
○​ Creative jobs may be affected by automated content generation.
ER

●​ Privacy Risk
LV

○​ Generative AI may generate sensitive personal information.


○​ This includes data like social security numbers or medical records.
SI

○​ Such information can be misused for harmful or malicious purposes.

Responsible Use of Generative AI


●​ Ensuring that the training data used are diverse and representative.
●​ The outputs are scrutinized for bias and misinformation.
●​ Prioritizing user privacy and consent
●​ Having clear guidelines around ownership and attribution of generative content.
●​ Discuss the social and ethical impact of Generative AI.
●​ Use Generative AI responsibly for society’s benefit.
●​ Focus on ethics to build trust in AI.
●​ Reduce negative effects through proper rules and safeguards.
●​ Support laws and guidelines for safe AI use.
●​ Encourage innovation to maximize AI’s positive impact on society.

L
O
O
H
SC
E
G
TI
ES
PR
E
N
LI
ER
LV
SI

You might also like