Scribd
Upload
6 views2 pages

Generative AI

Generative AI is a type of artificial intelligence that creates new content such as text, images, audio, video, and structured data by learning patterns from large datasets. It utilizes various model types like LLMs, diffusion models, and GANs, and is characterized by its creativity, context-awareness, and scalability. However, it also has limitations including the potential for hallucination, data bias, and the need for human oversight.

Uploaded by

Prakash Bhusal
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
6 views2 pages

Generative AI

Generative AI is a type of artificial intelligence that creates new content such as text, images, audio, video, and structured data by learning patterns from large datasets. It utilizes various model types like LLMs, diffusion models, and GANs, and is characterized by its creativity, context-awareness, and scalability. However, it also has limitations including the potential for hallucination, data bias, and the need for human oversight.

Uploaded by

Prakash Bhusal
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

Generative AI refers to a class of artificial intelligence systems that can create new

content—rather than just analyze or classify existing data.

What it can generate

●​ Text: answers, summaries, emails, code (e.g., ChatGPT)​

●​ Images: artwork, designs, realistic photos (e.g., DALL·E, Midjourney)​

●​ Audio: speech, music, sound effects​

●​ Video: short clips, animations​

●​ Structured data: synthetic datasets, test cases​

How it works (high level)

Generative AI models learn patterns from large amounts of data and then predict what
should come next.

●​ Common model types:​

○​ LLMs (Large Language Models) → text & code​

○​ Diffusion models → images & video​

○​ GANs / VAEs → images, data generation​

●​ Training involves:​

○​ Massive datasets​

○​ Neural networks (Transformers are dominant)​

○​ Fine-tuning and alignment for usefulness & safety​

Key characteristics

●​ Creative: produces novel outputs​

●​ Context-aware: adapts to prompts and instructions​

●​ Probabilistic: outputs may vary for the same input​


●​ Scalable: works across many domains with one model​

Real-world use cases

●​ Hiring & HR: interview questions, candidate evaluation, resume screening (relevant
to AI recruitment bots like the one you’re building)​

●​ Banking & Finance: report generation, customer support, risk summaries​

●​ Healthcare: clinical notes, medical summaries​

●​ Software: code generation, testing, documentation​

●​ Manufacturing & Quality: deviation analysis, CAPA drafting, change control


documentation​

Limitations & risks

●​ Can hallucinate (confident but incorrect outputs)​

●​ Data bias → biased outputs​

●​ Needs human oversight for critical decisions​

●​ Privacy and compliance concerns

You might also like