Generative AI is a subset of artificial intelligence that uses machine learning models to generate new content, such as text, images, video, and audio, in response to prompts or inputs. It is called generative because it can create new content that it has not seen or experienced before, rather than simply recognizing or classifying existing content.
One example of generative AI is a program called Perplexity AI, which is similar to other conversational AI models like ChatGPT or Bard, but with a research focus. Perplexity AI cites the sources it uses, making it a useful tool for researchers and students. Another example is Soundraw, a Tokyo-based company that uses generative AI to generate audio for musicians and creators. Generative AI models, such as those used by Perplexity AI and Soundraw, can increase worker productivity and creativity by generating new content on demand.
Generative AI models, such as those used by ChatGPT and DALL-E, start by encoding a representation of what they want to generate. For example, a generative AI model for text might represent words as vectors that characterize the similarity between words or their meaning. These models can then generate new content by sampling from the distribution of possible outputs, based on the input prompt and the encoded representation. Generative AI is a rapidly evolving field, with many potential applications and benefits, as well as ethical and societal considerations.