AI Models for Different Uses
AI Models for Different Uses
Learn about foundation, general-purpose, and task-specific AI tools and how each one helps in different ways.
Learning Objectives
By the end of this lesson, you will be able to:
- Describe the differences between foundation models, general-purpose models, and task-specific AI tools
- Match real-world AI tools (like ChatGPT, Grammarly, or Gemini) to their appropriate category
- Recognize which type of model is most appropriate for different kinds of tasks
- Understand how different AI tools are built and why that matters for ethical use
Take a Guess
Many assistive tools and technology use artificial intelligence (AI.)
Try one of these activities to take a guess about what AT uses AI today.
This activity is appropriate for learners needing movement and manipulation.
This activity is appropriate for learners using screen readers or needing traditional questions.
Think of AI like a pyramid with three main layers. The base are the foundation models. The middle are general purpose models, and the top are the specialized tools for particular jobs. People are building new models every day. Below are some examples.
Foundation Models
Foundation models are large AI systems that are trained on massive amounts of data. These foundations act like a brain that understand language, images and patterns. They require significant computing resources to build. Foundation models are created by major tech companies and research labs.
- OpenAI’s GPT models.
- Google’s Gemini.
- Meta’s LLaMA models.
General Purpose Models
General purpose models are built on top on foundation models but optimized for specific uses. They can handle conversations, writing, analysis, and various tasks. They are more accessible and practical for everyday use. These models are often available through user-friendly interfaces.
ChatGPT
Company – OpenAI.
Focus – Conversation and Creativity
Claude
Company – Anthropic.
Focus: Safe, ethical and conversational
GPT 4.0
Company – OpenAI.
Focus – Advanced reasoning and text generation
Co-Pilot
Company – Microsoft (GitHub).
Focus – Document Writing, data analysis, presentations
Gemini
Company – Google
Focus – More advanced foundation model that integrates text, images and potentially other modalities (replaces Bard)
Task Specific Models
Task specific models are specialized tools for particular jobs. They focus on doing one thing very well. Often, they are used by specific industries or for specific purposes.
Grammarly
Company – Owned by: Grammarly Inc. Grammarly is a private company founded by Max Lytvyn, Alex Shevchenko, and Dmytro Lider. It is based in San Francisco, California.
Pictory
Owned by: Pictory Inc. Pictory is a private company providing AI-powered video creation and editing tools. It was founded by Vikram Chalana and is headquartered in Bellevue, Washington.
Gamma
Owned by: Gamma Technologies Inc. Gamma is a presentation and storytelling tool powered by AI. It is a relatively new startup and operates as an independent entity.
Imagen
Imagen is an AI model developed by Google Research, specifically by the Brain Team at Google. It is not owned by a separate company, as it is a project under Google’s umbrella. Imagen is a text-to-image generative AI model designed to create high-quality images from descriptive text prompts.
Multi-Bot Platform
There is an application that allows you to compare outputs from different AI models. Poe.com is a website created by Quora that lets you chat with different AI bots, like ChatGPT and Claude, all in one place. One of the main benefits is that you can compare answers from multiple bots at the same time. There are free versions as well as a subscription plans.