AI Glossary
Essential AI terms every professional should know
Master the language of AI with our comprehensive glossary. From foundational concepts to cutting-edge techniques, understand the terminology shaping the future of work.
62 terms found
AI
Artificial Intelligence - The simulation of human intelligence processes by machines, including learning, reasoning, and self-correction.
Machine Learning
A subset of AI that enables systems to learn and improve from experience without being explicitly programmed.
Deep Learning
Advanced machine learning using neural networks with multiple layers to analyze data with increasing levels of abstraction.
Neural Network
Computing systems inspired by biological neural networks, consisting of interconnected nodes that process information.
LLM
Large Language Model - AI models trained on vast text datasets to understand and generate human-like text (e.g., GPT, Claude).
GPT
Generative Pre-trained Transformer - A type of LLM developed by OpenAI that generates human-like text.
Prompt
Instructions or questions given to an AI model to generate specific outputs or responses.
Prompt Engineering
The practice of designing effective prompts to optimize AI model outputs and achieve desired results.
Fine-tuning
The process of adapting a pre-trained model to perform specific tasks by training it on specialized datasets.
Training Data
The dataset used to teach an AI model patterns, relationships, and knowledge during the learning process.
Algorithm
A set of step-by-step instructions or rules that a computer follows to solve problems or perform tasks.
NLP
Natural Language Processing - AI technology that enables machines to understand, interpret, and generate human language.
Computer Vision
AI field focused on enabling machines to interpret and understand visual information from images and videos.
Generative AI
AI systems that create new content (text, images, music, code) rather than just analyzing existing data.
AGI
Artificial General Intelligence - Hypothetical AI with human-level intelligence across all cognitive tasks (not yet achieved).
Narrow AI
AI designed to perform specific tasks (like facial recognition or language translation), as opposed to AGI.
Supervised Learning
Machine learning where models learn from labeled training data with known correct answers.
Unsupervised Learning
Machine learning where models find patterns in unlabeled data without predefined categories.
Reinforcement Learning
Machine learning where agents learn optimal behavior through trial, error, and rewards.
Token
Basic unit of text that LLMs process (can be a word, part of a word, or punctuation).
Context Window
The maximum amount of text an AI model can consider at once when generating responses.
Hallucination
When AI generates false or nonsensical information presented as fact, often occurring when uncertain.
Bias
Systematic errors in AI outputs caused by prejudiced training data or flawed algorithms.
Overfitting
When a model learns training data too precisely, including noise, reducing its ability to generalize.
Underfitting
When a model is too simple to capture underlying patterns in data, resulting in poor performance.
Embedding
Mathematical representation of words or concepts as vectors, capturing semantic relationships.
RAG
Retrieval-Augmented Generation - Combining AI generation with external knowledge retrieval for accurate responses.
API
Application Programming Interface - A way for different software applications to communicate and share data.
Transfer Learning
Using knowledge gained from one task to improve learning and performance on related tasks.
Transformer
Neural network architecture that revolutionized NLP by using attention mechanisms (foundation of GPT, BERT).
Attention Mechanism
AI technique that helps models focus on relevant parts of input data when making predictions.
Zero-Shot Learning
AI's ability to perform tasks it wasn't explicitly trained for, using general knowledge.
Few-Shot Learning
AI performing tasks with only a few examples, rather than extensive training data.
Chain-of-Thought
Prompting technique where AI shows step-by-step reasoning to arrive at conclusions.
Temperature
Parameter controlling randomness in AI outputs (lower = more focused, higher = more creative).
Inference
The process of using a trained AI model to make predictions or generate outputs on new data.
Model
The trained algorithm and parameters that enable AI to make predictions or generate outputs.
Parameter
Adjustable values in a model that are learned during training to make accurate predictions.
Epoch
One complete pass through the entire training dataset during the learning process.
Batch Size
The number of training examples processed together before updating the model.
Learning Rate
How quickly a model adjusts its parameters during training (too high = unstable, too low = slow).
Activation Function
Mathematical function that determines whether a neuron should activate based on input.
Loss Function
Metric measuring the difference between predicted and actual outputs during training.
Optimizer
Algorithm that adjusts model parameters to minimize errors and improve performance.
Gradient Descent
Optimization algorithm that iteratively adjusts parameters to find the best model configuration.
Backpropagation
Method for calculating gradients and updating neural network weights during training.
Convolutional Neural Network
Deep learning architecture specialized for processing grid-like data (images, videos).
Recurrent Neural Network
Neural network designed to process sequential data by maintaining memory of previous inputs.
Autoencoder
Neural network that compresses data into lower dimensions and then reconstructs it.
GAN
Generative Adversarial Network - Two AI models competing: one generates fake data, the other detects fakes.
Diffusion Model
AI model that generates images by gradually removing noise from random data.
Stable Diffusion
Open-source diffusion model for generating images from text descriptions.
DALL-E
OpenAI's image generation model that creates images from textual descriptions.
Midjourney
Popular AI image generation platform known for artistic and stylized outputs.
Sentiment Analysis
NLP technique that identifies emotional tone in text (positive, negative, neutral).
Tokenization
Breaking text into smaller units (tokens) that AI models can process.
Classification
Task of categorizing data into predefined classes or labels.
Regression
Predicting continuous numerical values based on input data.
Clustering
Grouping similar data points together without predefined categories.
Explainable AI
AI systems designed to provide understandable explanations for their decisions and outputs.
Edge AI
Running AI models on local devices (phones, IoT) rather than cloud servers for faster, private processing.
Synthetic Data
Artificially generated data used to train AI models when real data is scarce or sensitive.
What's Your AI Survival Type?
Take our 3-minute test to discover your position and strategy in the AI era.
Start Test