Fundamentals of ML and AI
Introduction
AI, ML, deep learning, and generative AI are nested fields. Each one is a subset of the one before it.
| Field | What it is | Example |
|---|---|---|
| Artificial intelligence (AI) | Umbrella term for systems that do tasks needing human intelligence: perception, reasoning, learning, problem solving, decision making | Any intelligent system |
| Machine learning (ML) | A type of AI where machines learn from data to improve performance on a task | Fraud prediction model |
| Deep learning (DL) | ML that uses layered neural networks, modeled on neurons and synapses | Amazon Rekognition analyzing images and video |
| Generative AI | A subset of deep learning that generates new content from patterns learned in training data | Text, images, audio, code |
Key points
- Traditional AI/ML analyzes and interprets data. Generative AI creates new content.
- Generative AI can adapt models built with deep learning without retraining or fine-tuning.
- Example models: Amazon Titan and Anthropic Claude for text, Stable Diffusion for images.
- Uses include text generation, image generation, speech synthesis, and code generation.
- Generative AI raises responsibility concerns such as bias, privacy, and responsible use.