It’s AI that can create new content, such as text, images, video, music, and speech. Examples of models that generate text include ChatGPT, Microsoft Copilot, and Gemini. For generating images, tools like MidJourney, Adobe Firefly, and Stable Diffusion are popular. In the realm of speech generation, Eleven Labs is widely used, while tools like Suno and Udio specialize in generating music.
Another type of AI is “discriminative AI,” which classifies, predicts, or recognizes patterns in existing data. Examples include Netflix's recommendations for what to watch next and Gmail's spam filtering.
It’s useful to differentiate between these two types of AI when discussing their applications. Is it classifying existing data (as with spam filtering), or generating new content (as with ChatGPT)? These systems have distinct strengths and weaknesses based on their design and purpose.
Adapted from "FAQs about generative AI" by Nicole Hennig, University of Arizona Libraries. Licensed under CC BY 4.0.
