Introduction to Generative AI

Instructor

• Dr. Gwendolyn Stripling
• Artificial Intelligence Technical Curriculum Developer at Google Cloud

Course Overview

• Define generative AI
• Explain how generative AI works
• Describe generative AI model types
• Describe generative AI applications

What is Generative AI?

• A type of AI technology that can produce content like text, imagery, audio, and synthetic data.

Understanding AI and Machine Learning

• Artificial Intelligence (AI): A discipline aimed at creating intelligent agents capable of reasoning, learning, and acting autonomously.
• Machine Learning (ML): A subset of AI, involving training models from input data to make predictions on new data without explicit programming.
  • Supervised ML: Models trained on labeled data to predict outcomes.
  • Unsupervised ML: Models trained on unlabeled data to discover patterns or groupings.
  • Deep Learning: Uses neural networks to process complex patterns and can work with both labeled and unlabeled data (semi-supervised learning).

Generative AI

• A subset of deep learning using neural networks to create new content.
• Works with both labeled and unlabeled data using supervised, unsupervised, and semi-supervised methods.

Types of AI Models

• Discriminative Models: Classify or predict labels for data points (e.g., distinguishing a dog from a cat).
• Generative Models: Learn from existing data to generate new data instances (e.g., generating an image of a dog).

Characteristics of Generative AI

• Generates new content based on learned patterns from existing data.
• Utilizes transformers, which include encoder-decoder structures.
• Handles various forms of input and output, including text, images, audio, and video.

Applications of Generative AI

• Language Models: Generate natural language text.
• Image Models: Create images based on text descriptions.
• Video and 3D Models: Generate videos or 3D objects from text.
• Text-to-Task Models: Perform actions or tasks based on text inputs.

Foundation Models

• Large AI models pre-trained on extensive data for adaptation to various tasks.
• Examples include sentiment analysis, image captioning, and object recognition.

Google's Generative AI Tools

• Generative AI Studio: Tools for exploring and customizing GenAI models.
• Generative AI App Builder: Allows creation of GenAI applications without coding.
• Palm API and Maker Suite: Integrate with tools for model training, deployment, and monitoring.

Example Use Cases

• Code Generation: Convert code, debug, explain, craft queries, translate, and generate documentation.
• Generative AI Applications: Create digital assistants, custom search engines, and knowledge bases.

Conclusion

• Generative AI leverages large language models to revolutionize content generation across various media types.

Thank you for participating in the "Introduction to Generative AI" course.