Glossary
The words, without the hand-waving.
Definitions of the AI and web data terms developers actually encounter, written to be read once and understood.
A
Agentic AIAgentic AI pursues complex goals with autonomy. Unlike a chatbot that waits for your next prompt, Agentic AI proactively plans, executes, and adapts until the job is done.Agentic WorkflowAgentic Workflow is an iterative system where an LLM doesn't just talk but acts like planning tasks, web search, code execution, data fetch, and self-correcting errors to achieve a complex goal without human intervention.AI AgentsAI agents are intelligent systems that are capable of reasoning, planning, deciding and executing multistep tasks to achieve a goal. AI GatewayAn AI Gateway is an intelligent intermediary for managing, securing, and routing interactions between applications and various AI models.AI HallucinationAI hallucination refers to when an AI model confidently generates false, nonsensical, or factually incorrect information or unsupported by evidence. AI InferenceAI Inference is the process of using a trained AI model to make predictions or decisions on new, unseen data. It is the phase of AI when models stop learning and start producing real-world resultsAI TokensAI tokens are the fundamental units of text that LLMs process to understand and generate text. They are the building blocks that enable AI to interpret prompts and create responses.AlgorithmA clear and finite set of step-by-step instructions used by computers to solve problems, make decisions, or process data is called an algorithm.Artificial General Intelligence (AGI)Artificial General Intelligence (AGI) is hypothetical AI that can match human-level intelligence to understand, learn, and perform any intellectual task like a human being, rather than being designed or trained for a single, narrow-focused task.Artificial IntelligenceArtificial Intelligence (AI) is the capability of machines to perform tasks that typically require human intelligence, like learning and reasoning.Artificial Narrow Intelligence (ANI)Artificial Narrow Intelligence (ANI), also known as Weak AI, specializes in performing one or a limited set of tasks with high efficiency and precision, but lacks general human‑level thinking.Artificial Superintelligence (ASI)Artificial Superintelligence (ASI) is hypothetical AI exceeding human intelligence in every intellectual domain, including creativity and strategy.
C
Chain of ThoughtChain-of-Thought (CoT) is a reasoning approach that instructs AI models to think step-by-step and generate intermediate results before arriving at the final answer. ChatGPTChatGPT is OpenAI’s conversational interface that allows users to interact with its LLMs to generate text, code, and analysis.Claude AIClaude is a family of large language models (LLMs) developed by Anthropic, an AI safety-focused company. Cognitive ComputingCognitive computing is a branch of artificial intelligence that simulates human thought processes to solve complex problems and make better decisions.Context WindowA context window is the maximum number of tokens an LLM can process during a single interaction to maintain conversation history and generate coherent responses.Convolutional Neural Network (CNN)A CNN is a deep learning model that processes and classifies images by detecting patterns like edges, shapes, and objects.Cursor AICursor is an AI-powered code editor that includes advanced AI features like autocomplete, chat, and agent-based automation directly into the coding experience.
D
G
GeminiGemini is Google's native multimodal Generative AI assistant for handling text, images, video, audio, and code in conversations via web and apps.Generative Adversarial Network (GAN)Generative Adversarial Networks (GANs) are AI frameworks where two neural networks are pitted against each other to create realistic images, audio, and other synthetic data. Generative AI (GenAI)Generative AI is part of Artificial Intelligence that doesn't just analyze existing data, but it creates new data. It can generate text, images, code, audio, and video that has never existed before.GPTGenerative Pre-trained Transformers (GPT) are a family of large language models developed by OpenAI, capable of generating human-like text.GPU (Graphics Processing Unit)A GPU is a specialized processor designed to perform thousands of calculations simultaneously, initially created for rendering graphics but now crucial for AI.GrokGrok is xAI’s chatbot, designed to deliver direct, often witty responses, with real-time access to public data from the X platform. It is built with fewer content filters and a more candid tone compared to other AI assistants.
M
Machine Learning (ML)Machine learning is a science of getting computer systems to act without being explicitly programmed. Instead of writing rules, you feed data to an algorithm, and it figures out the rules itself.Manus AIManus AI is an autonomous general-purpose AI agent that executes complex, multi-step tasks on a chat interface similar to ChatGPT and other LLMs. Mistral AIMistral AI is a Paris-based startup founded in 2023, specializing in efficient, open-weight LLMs. Model Context Protocol (MCP)MCP is an open standard created by Anthropic that lets AI models connect directly to your data sources and tools without custom integrations.Multimodal AIMultimodal AI is a model that can process and reason across multiple data types like text, images, audio, and video.
N
Natural Language Processing (NLP)Natural Language Processing (NLP) is a field of artificial intelligence that enables computers to process, understand, and generate human language. Neural NetworksA machine learning model inspired by the human brain that learns patterns by passing data through interconnected layers of artificial neurons.
P
Perplexity AIPerplexity AI is an answer engine that combines capabilities of large language models with real-time web search to give concise responses with citations.PromptPrompts are the instructions you give an AI model to generate answers, blog posts, code, summaries, or images, as required.Prompt EngineeringPrompt engineering is the practice of designing and refining prompts to guide AI models toward accurate, and desired results.
R
Recurrent Neural Network (RNN)Recurrent Neural Networks (RNNs) are neural networks that process sequential data by maintaining a memory of previous inputs and are ideal for tasks like language modeling and time-series forecasting.Reinforcement LearningReinforcement learning (RL) is a machine learning method where an AI agent learns through trial and error, receiving rewards and penalties, to make better decisions over time.Retrieval-Augmented Generation (RAG)RAG is a technique that gives an LLM a reference library. It retrieves specific data from your documents and feeds it to the AI so it can answer questions accurately without hallucinating.
S
Sora AISora is an AI model developed by OpenAI (the creators of ChatGPT) that can create realistic and imaginative videos from text instructions.Supervised LearningSupervised learning is a machine learning method where models are trained using labeled data so it can predict correct outputs on new data.
T
Training DataTraining Data is the raw information AI learns from to recognize patterns, make predictions, or perform specific tasks.TransformerTransformers are a neural network architecture that uses self-attention to process sequential data that powers language, vision, and multimodal AI models.Turing TestThe Turing Test is a method proposed by Alan Turing to determine if a machine is capable of carrying on a conversation and exhibiting intelligence like a human without being identified as a machine.
Reading about it is the slow way.
500 free credits a month, no card, every endpoint unlocked. Try the APIs behind half of these terms.