Google Gemini, Microsoft CoPilot, ChatGPT, and others are examples of "generative Artificial Intelligence" or GenAI. These tools differ from a Google search - where you are given results to a specific search inquiry. Instead, GenAI tools create “new content” by predicting the word most likely to come next (e.g. based on publicly available Internet sites including Wikipedia and YouTube transcripts).
GenAI programs are designed to model and imitate human language. They use mathematical models to predict what the next word is most likely to be based on what you are asking for, but because they use natural language, it "feels" like they are thinking. GenAI programs are trained on huge datasets of publicly available Internet sites including places like Wikipedia and Youtube transcripts. However, evaluating the quality of the sources is not a part of the GenAI output.
The field of AI is changing at a rapid pace. We know that these generative tools help users synthesize information and create content (code, essays, art, music, etc.). However, these tools can also "hallucinate", or make up facts or sources and create biased content.
Content adapted from University of Minnestota Library and UC San Diego Library
Image text and description:
Defining Generative AI
To understand generative artificial intelligence (GenAI), we first need to understand how the technology builds from each of the AI subcategories listed below.
Artificial Intelligence, the broadest category, is the theory and methods to build machines that think and act like humans (shows an illustration of a brain with brightly colored areas). Programmers teach AI exactly how to solve specific problems by providing precise instructions and steps.
Machine Learning, a subcategory of Artificial Intelligence, is the ability for computers to learn from experience or data without human programming (shows an illustration of a laptop with a lightbulb on the screen).
Deep Learning, a subcategory of Machine Learning, mimics the human brain using artificial neural networks such as transformers to allow computers to perform complex tasks (shows an illustration of a network with nodes and connections).
Generative AI, a subcategory of Generative AI, generates new text, audio, images, video or code based on content it has been pre-trained on (shows decorative icons beside ChatGPT, Midjourney, and Bard).
Content adapted from the University of North Carolina Library