Skip to Main Content

Generative Artificial Intelligence (GenAI)

Types of Generative AI

  1. Text: AI tools that generate human-like text based on prompts (i.e. ChatGPT)
  2. Images: Generative Adversarial Networks (GANs) can create images of objects, scenes, and people that do not actually exist. They can also enhance image resolution, create art, or design virtual environments. (i.e. Midjourney)
  3. Audio: AI can generate music, sound effects, and voice audio, mimic real instruments, or create new instruments. Text-to-Speech (TTS) systems convert text into spoken word in a natural-sounding voice, while Speech-to-Text (STT) systems convert audio tracks into written words. (i.e. Elevenlabs)
  4. Video: AI generates video clips by creating new frames or altering existing footage. This can be used to create visual effects, "deep fakes", or animated features. (i.e. Sora)
  5. Synthetic Datasets: AI can create datasets that mimic real-world data, which then can be used for training machine learning models. This is helpful for times when real-world data contains private or sensitive information, and should not be input into an AI tool for training. (i.e. Mostly)
  6. 3D Models: AI can design 3D models which can be used for various purposes in entertainment, architecture, and other industries. (i.e. 3Dfy AI)
  7. Code: AI can assist in generating code for computer programming, or aid software development by converting natural language requirements into code structures. (i.e. GitHub Copilot)

 

References

Shirer, P. (2023). What are the key types of generative AI content? LinkedIn. https://www.linkedin.com/pulse/what-key-types-generative-ai-content-paul-shirer-tyw5c/