Copilot AI Glossary 🚀
New terms and acronyms in the growing AI World
Microsoft Copilot is your everyday AI companion, providing accessible AI capabilities based on your needs and preferences!
Artificial Intelligence or AI — the simulation of human intelligence in machines that are programmed to think and learn like humans.
Algorithm — is a set of rules or instructions given to an AI system to help it learn from data.
AI copilot — is a conversational interface that uses large language models (LLMs) to support users in various tasks and decision-making processes across multiple domains within an enterprise environment.
Copilot — Copilots are natural language assistants that can help with creative tasks, generate insights, execute automated workflows, and more. Copilots are composed of workflows, actions, knowledge, and triggers, powered by one or more foundation models and an orchestrator that oversees and synchronizes the operations of the copilot.
Custom Copilot — is a custom version of Microsoft Copilot that combines instructions, extra and/or custom knowledge, and any combination of skills.
Microsoft Copilot — an accessible, cohesive AI interface that provides users with access to AI capabilities based on their needs and preferences, while also integrating with Microsoft products to maximize value. Microsoft Copilot is your everyday AI companion.
Microsoft Copilot Studio — A low/no code tool that allows users to easily integrate artificial intelligence into any M365 or Power Platform product, offering prebuilt and custom AI models and systems for tasks like form processing, object detection, prediction and more.
Copilot extensions — a copilot extension customizes and enhances Microsoft Copilot with custom copilots, enabling new actions and customized knowledge for grounding within Copilot. With Copilot extensions, users can get a Microsoft Copilot experience that is tailored with the data, systems and workflows they use every day.
ChatGPT — A chat interface built on top of GPT-3.5. GPT-3.5 is a large language model developed by OpenAI that is trained on a massive amount of internet text data and fine-tuned to perform a wide range of natural language tasks.
Chatbot — A user-friendly interface that allows the user to ask questions and receive answers.
Generative AI — a form of AI characterized by its ability to create natural language/more human-like content suggested by input prompts, including prose, verse, music, and images. GPT — (generative pretrained transformer) a class of foundation models created by OpenAI and hosted by OpenAI and Azure. A recent model in this class is “GPT-4 Turbo”.
Grounding — is the process of linking abstract knowledge in AI systems to specific, real-world content. It increases the accuracy of AI agent comprehension of and interaction with real world data.
LLMs (Large Language Models) — generative AI models that are trained on a massive trove of data to produce human-like responses to natural language queries, typically through a chatbot. See also: Foundation model.
LLMOps — streamlined flow for end-to-end development of LLM-powered applications from ideation to operationalization.
Low-Code — typically involves graphical/visual interfaces and minimal coding to allow rapid, accessible application development. Unlike pro-code tools, most if not all underlying concepts and technologies are abstracted away from the user experience.
Prompt — the input to a generative AI model from which it generates an output (often called an “answer” or “completion”). Usually text, but multimodal models can use text, images, audio, or a combination of these as the prompt.
Plugins — a type of copilot extension. Microsoft has defined a new plugin manifest that unlocks the ability to write a plugin once and run it anywhere on any copilot surface. Plugins should be considered an atomic, functional extensibility artifact that can be composed with any other copilot extension.
Responsible AI (RAI) — is the set of norms and standards that Microsoft seeks to define to help advance the safe and secure use of AI for the benefit of society at large through governance, internal policy, enablement, external engagement, and thought leadership.
Tokenization — the process of breaking text into individual words or subwords to input them into a language model.
Transcribing — is part of Speech recognition, wich involves converting speech into a text representation.
Teams Message Extension — a feature of Microsoft Teams that allows users to search or initiate actions in a web service / external system through a simple UX element called an Adaptive Card. These are all now usable as plugins.
Whisper — OpenAI’s Whisper is an AI system developed to perform automatic speech recognition (ASR), the task of transcribing spoken language into text.
I pulled the majority of content from Microsoft Docs and Learn.
I hope you find this post helpful in your AI journey! You can add your new AI terms in the comments.
Thank you for reading! 😊
