generative ai Flashcards
what is generative ai?
1) generative artificial intelligence (ai) refers to algorithms that generate new data that resembles human generated content, such as audio ,code ,images ,text, videos and stimulations
2)this technology is trained with existing data and content which creates the potential for applications such as natural language processing, computer vision, metaverse and speech synthesis
generative ai vs conventional ai?
goal: generate ai creates new content and conventional ai analyzes , processes and classifies existing data
training : ai generative models use vast libraries of sample to rain neural networks and other complicated structures to create new content based on existing content. Conventional ai uses fewer complex.
output : generative ai outputs are innovative, , new and often unexpected. Conventional ai output is predictable based on existing data
applications: generative ai benefits art, music , literature , gaming , design
conventional ai is used in banking , healthcare , image recognition and language processing.
language models
trained to give output based on the prompt entered by the user
examples: BERT (google)
T5 (google)
GPT - 3 & GPT - 4 (OpenAI’s)
Chatbots
trained to stimulate conversation with the user
examples: ChatGPT(OpenAI’s)
Google’s Meena
GANs
(Generative adversarial Networks )
generates realistic images from random noise
Example : StyleGAN & BigGAN
VAEs
(variational autoencoders )
generates images based on distribution of input data
Diffusion models
refines noise to generate images
example: DALL-E2
Stable diffusion
music generation
ai that generates music
example: Jukedeck and MuseNet
speech synthesis
ai that converts text into natural flowing speech
examples : google WaveNet
Amazon Polly
deepfakes
ai that generates videos that can superimpose faces
example: DeepFaceLab & FaceApp
generative video model
ai that generates video content
examples: VQ-VAE
MoCoGAN
3D object generator
3DGEN
pointNET
code generator
ai that generates or autocompletes code
example: codex & copilot
what are the benefits of using generative ai?
1) creativity : generative ai can help inspire people. it will be very beneficial in fields such as art, design ,music etc
2) efficiency : generative ai can automate content creation processes which saves time and reduces cost compared to the normal traditional methods
3) personalization : generative ai can created content based on user behavior and preferences such as product recommendation or personalized news letters
4) exploration : generative ai can help explore new design spaces optimize complex systems such as developing a new drug or improving industrial processes
5) accessibility : generative ai improves access to resources making it easier for people who have limited sources.
limitations of generative ai?
1)ethical concerns : the creation and spread of fake content, such as deep fake videos, fake news articles ,forged document etc can spread misleading information, manipulate people , identity theft which will not benefit the society
2) bias and fairness
generative ai inherits bias from its training data resulting in discriminatory outcomes
3) energy consumption : training large generative ai model requires a lot of energy and resources
4) cost of development : the cost and training of generative ai model can be extremely expensive.
5) data dependency : these models require a vast amount of data to generate content effectively.
