S
SurvTest

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
AI Glossary | AI Survival Test | AI Survival Test