Generative Artificial Intelligence

Many things that machines do fall under the term "artificial intelligence" (AI). When AI creates content, it is referred to as generative. How might this work?

The AI Landscape

AI has a long history.

The first use of the term Artificial Intelligence was at the famous 1956 Dartmouth Summer Research Project on Artificial Intelligence, organized by John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon. The term was created to unify topics in cybernetics, automata theory, and information processing related to:

Programs that would communicate in natural language
Programs that could form concepts and abstractions
Programs that could solve complex problems previously exclusive to humans
Programming via mimicking biological neurons
Programs that could learn and write new code to improve themselves

Gen AI did not begin in the 2020s.

Timeline

TODO

Fields of AI

TODO

Traditional versus Generative AI

Roughly, traditional AI is designed to analyze data, recognize patterns, make predictions, and automate tasks. Generative AI creates new, never before seen content (text, images, video, media, code).

Traditional AI	Generative AI
Analyzes data to recognize patterns	Creates new content
Rule-based (may use statistics though)	Deep learning-based
Does not require a lot of data	Requires massive datasets for training
Automates specific tasks	Can assist in creative processes
Processing, classifying, predicting, insight discovery	Creation and synthesis
Applications: fraud detection, spam filtering, movie recommendations	Applications: content generation, image synthesis, code generation, conversational assistants, coding agents.

Algorithms and Models

TODO

Infrastructure

TODO

Ethics, Alignment, and Safety

TODO

Recall Practice

Here are some questions useful for your spaced repetition learning. Many of the answers are not found on this page. Some will have popped up in lecture. Others will require you to do your own research.

TODO

Summary

We’ve covered:

TODO
TODO