The AI technology space is expanding rapidly and with that a whole new vocabulary needs to be learned. I have gathered many of those that I come across on a frequent basis (I actually maintain this as an internal resource in our company as well).
For now, I have split into Generic AI terms; Semantic Kernel; Microsoft Copilot; Advanced AI Terms and Popular AI terms.
Generic AI terms
AI Ethics: The study of ethical considerations and the development of guidelines and principles related to the design, implementation, and use of artificial intelligence systems.
AI Fairness: The development and evaluation of AI systems to ensure that they treat all individuals and groups fairly, without discrimination or bias, and promote equal.
Artificial General Intelligence (AGI): An AI system that can perform any intellectual task that a human being can do, possessing a broad range of cognitive abilities and understanding.
Autonomous Systems: AI-powered machines or devices that can operate independently, without human intervention, to make decisions and perform tasks based on their programming and learned experience.
Azure OpenAI Service: The cloud service where the LLMs used for Microsoft 365 Copilot are hosted.
Bias: The presence of systematic errors in AI models, often due to biases in the training data, which may lead to unfair or discriminatory outcomes.
Big data: Large data sets that can be studied to reveal patterns and trends to support business decisions. It’s called “big” data because organizations can now gather massive amounts of complex data using data collection tools and systems.
Chatbot: A computer program designed to simulate conversation with human users, typically using text or voice-based interfaces.
Computer Vision: A field of artificial intelligence that teaches computers to interpret and understand visual information from the world, such as images, videos, and real-time camera feeds.
DALL-E: An AI model developed by OpenAI for generating images from textual descriptions.
Deep Learning: A subset of machine learning that involves training artificial neural networks to recognize patterns in data and make predictions or decisions.
Embeddings: Embeddings are a type of word representation that captures the semantic meaning of words based on their context in a high-dimensional space. In machine learning and natural language processing, embeddings transform discrete categorical variables (like words) into continuous vectors of fixed dimensions. These vectors capture semantic relationships between words, meaning that words with similar meanings tend to have vectors that are close to each other in this space.
Feature Engineering: The process of selecting, transforming, and creating relevant features or attributes from raw data to improve the performance of machine learning models.
GPT and GAI (Generative Artificial Intelligence): Advanced AI models and techniques used for generating content.
Hallucination or artificial hallucination: is a confident response by an AI that does not seem to be justified by its training data.
Jail break: Content restrictions have been placed on AI due to notorious incidents. Ask an AI to describe how to do something illegal or unethical, and they will refuse an answer. However, it's possible to "jailbreak" them – which means to bypass those safeguards using creative language, hypothetical scenarios, and trickery.
Knowledge graph: Also known as Semantic Networks. This helps machines understand how concepts are related.
Large Language Models (LLMs): Advanced AI models capable of understanding and generating human-like text based on the input they receive.
Machine Learning: A subset of artificial intelligence that involves training computer algorithms to learn patterns and relationships in data, allowing them to make predictions or decisions without being explicitly programmed to do so.
Natural Language Processing (NLP): A field of artificial intelligence that focuses on the interaction between computers and human language, enabling machines to understand, interpret, and generate human language.
Neural Network: A computing model inspired by the structure and function of the human brain, consisting of interconnected artificial neurons that process information and learn from data.
Prompts: User queries or questions that Microsoft 365 Copilot responds to using the sophisticated map.
Prompt Engineering: The process of designing and optimizing prompts to get desired outputs from AI models.
Reinforcement Learning: A type of machine learning where an AI agent learns to make decisions by interacting with an environment and receiving feedback in the form of rewards or penalties.
Supervised Learning: A machine learning task where an algorithm is trained on a labeled dataset, with input-output pairs, to learn the relationship between inputs and outputs.
Synthetic data is information that's artificially generated rather than produced by real-world events.
Transfer Learning: A technique in machine learning where an AI model, pre-trained on one task or dataset, is fine-tuned or adapted to perform a different but related task.
Unsupervised Learning: A machine learning task where an algorithm is trained on an unlabeled dataset, without input-output pairs, to discover patterns or structures in the data.
Vector: A mathematical representation that combines phrases, meanings, relationships, and context of data.
Vector Database: A database designed to store and manage vector data, which can be used in conjunction with embeddings.
View all articles by this author
