alci.dev/en

Ai

38 terms

ABCDEFGHIJKLMNOPQRSTUVWXYZ

A

Agent A software entity that can take actions autonomously on behalf of a user.AI (Artificial Intelligence) The broad field of creating systems that can perform tasks that normally require human intelligence.Alignment (AI) The challenge of ensuring AI systems act in accordance with human values and intentions.Ambient AI AI that operates seamlessly in the background of environments and interactions without explicit user commands.Anthropomorphization The tendency to attribute human characteristics, emotions, or consciousness to AI systems.Automation (AI) Using artificial intelligence to automatically perform tasks that previously required human intervention or decision-making.AX (Agentic Experience) An extension of UX for the AI Age, designing agentic products that feel like relationships rather than tools.

B

Benchmark (AI) A standardized test or dataset used to evaluate and compare the performance of AI models across specific tasks.

C

Chain of Thought (CoT) A prompting technique that guides AI models to break down complex reasoning into step-by-step thought processes.Cluster (GPU Cluster) A collection of graphics processing units working together to train or run AI models at scale.Computer Use An AI capability that allows models to interact directly with computer interfaces like a human user would.Context (AI) The surrounding information and situational awareness that AI systems use to understand and respond appropriately.Context Window The amount of text or information an AI model can consider at one time when generating responses.Copilot (AI) An AI assistant that works alongside humans to enhance productivity and provide intelligent suggestions in real-time.Credits/Tokens (AI) Units of measurement and payment for AI API usage, representing computational resources consumed by AI operations.

D

Deterministic (AI) AI behavior that produces the same output given identical inputs, ensuring predictable and repeatable results.

E

Embeddings Numerical representations that capture the meaning and relationships between words, concepts, or data points in high-dimensional space.Escape Hatch A fallback mechanism that allows users or systems to override AI decisions or revert to manual control when needed.Evals Systematic evaluations and tests designed to measure AI model capabilities, safety, and performance across various tasks.Evaluation Harness An evaluation harness is a standardized AI testing framework for benchmarking LLM performance across tasks. Learn how tools like lm-eval-harness, HELM, and custom harnesses work.Explainability The ability of AI systems to provide clear, understandable explanations for their decisions and reasoning processes.

F

Feedback Loop A process where AI systems use outputs or user responses to improve future performance and decision-making.Few-Shot Learning An AI approach where models learn to perform new tasks using only a small number of training examples.

G

Generative AI AI systems that create new content like text, images, code, or audio based on patterns learned from training data.Generative UI AI-powered interfaces that dynamically create and adapt user interface elements based on context and user needs.GPT Generative Pre-trained Transformer, a type of large language model architecture that generates human-like text.Ground Truth The accurate, verified, or correct answer that serves as the benchmark for training and evaluating AI models.Grounding Connecting AI-generated content to factual sources and real-world knowledge to reduce hallucinations and improve accuracy.Guardrails Safety mechanisms and constraints that prevent AI systems from generating harmful, inappropriate, or unwanted outputs.

H

Hallucination When AI systems generate plausible-sounding but factually incorrect or fabricated information not based on training data.Human-in-the-Loop (HITL) An AI approach that incorporates human judgment and oversight at critical decision points in automated processes.

I

Inference The process of using a trained AI model to make predictions or generate outputs from new, previously unseen input data.Instruction-Following Model An instruction-following model is an AI system trained to understand and execute natural language commands. Complete guide with training methods, examples, RLHF, and applications.

L

Large Language Model (LLM) An AI model trained on vast text data to understand and generate human-like language.

M

Machine Learning (ML) The broader field of training algorithms to improve from data without explicit programming.

P

Prompt Engineering The practice of designing effective prompts to guide AI behavior and improve output.

R

RAG (Retrieval-Augmented Generation) A technique that enhances large language model outputs by incorporating information from external knowledge sources.

V

Vibe Coding A programming approach where developers describe what they want in natural language and let AI generate the code.