Week 5AI Fundamentals
AI Glossary: The Complete Bilingual Reference — English & Arabic 2026
30+ essential AI terms defined in both English and Arabic. Covers LLMs, RAG, AI Agents, Prompt Engineering, Fine-Tuning, Embeddings, Hallucination, MCP, and more — your go-to reference for AI terminology in the GCC.
Your Complete AI Glossary — Bilingual English & Arabic
This reference guide covers 30+ essential AI terms used daily in business, technology, and marketing contexts — with definitions in both English and Arabic. Bookmark it, share it with your team, and use it to decode the AI conversation.
Core Concepts
| Term (EN) | Arabic | Definition |
|---|
| AI — Artificial Intelligence | الذكاء الاصطناعي | The field of computer science focused on building machines that can perform tasks that typically require human intelligence, such as reasoning, learning, and problem-solving. |
| LLM — Large Language Model | نموذج اللغة الكبير | A type of AI model trained on enormous amounts of text data to understand and generate human language. Examples: GPT-4, Claude, Gemini. |
| Generative AI | الذكاء الاصطناعي التوليدي | AI that can create new content — text, images, audio, video, or code — based on patterns learned during training. |
| Machine Learning (ML) | التعلم الآلي | A branch of AI where systems learn from data and improve their performance over time without being explicitly programmed. |
| Deep Learning | التعلم العميق | A subset of machine learning that uses multi-layered neural networks to model complex patterns in data, especially unstructured data like images and text. |
Model Architecture & Behaviour
| Term (EN) | Arabic | Definition |
|---|
| Neural Network | الشبكة العصبية الاصطناعية | A computing system loosely inspired by the human brain, made up of interconnected nodes (neurons) that process information in layers. |
| Transformer | المحوّل | The neural network architecture behind most modern LLMs. It uses an 'attention mechanism' to understand relationships between words across long texts. |
| NLP — Natural Language Processing | معالجة اللغة الطبيعية | A field of AI focused on enabling computers to understand, interpret, and generate human language in a meaningful way. |
| Prompt | المطالبة / الموجِّه | The instruction or question you give to an AI model. A well-crafted prompt leads to better, more accurate responses. |
| Prompt Engineering | هندسة المطالبات | The practice of designing and refining prompts to guide AI models toward more accurate, relevant, or useful outputs. |
| Token | الرمز المميز / التوكن | The basic unit of text that AI models process. A token is roughly 3/4 of a word. Models have limits on how many tokens they can handle in one conversation. |
| Context Window | نافذة السياق | The maximum amount of text (in tokens) that a model can 'see' and consider at one time during a conversation. Larger context windows allow for longer, more complex conversations. |
| Hallucination | الهلوسة (في الذكاء الاصطناعي) | When an AI model generates information that sounds confident but is factually incorrect or fabricated. A key limitation to be aware of when using LLMs. |
| Temperature | درجة الحرارة (في النماذج) | A setting that controls how creative or random an AI's responses are. Low temperature = more predictable, focused answers. High temperature = more creative, varied responses. |
| Zero-shot | التعلم بدون أمثلة (زيرو-شوت) | When you ask a model to perform a task with no examples. The model relies entirely on its pre-trained knowledge to complete the task. |
| Few-shot | التعلم بأمثلة قليلة (فيو-شوت) | When you provide a small number of examples in the prompt to help the model understand what output format or style you want. |
| Chain of Thought (CoT) | سلسلة التفكير | A prompting technique where you ask the model to reason step-by-step before giving a final answer, leading to more accurate results on complex tasks. |
Training & Fine-Tuning
| Term (EN) | Arabic | Definition |
|---|
| Training | التدريب | The process of exposing an AI model to large datasets so it can learn patterns, relationships, and knowledge. Training happens before deployment. |
| Fine-Tuning | الضبط الدقيق | Additional training of a pre-trained model on a specific, smaller dataset to make it better at a particular task or domain (e.g., medical advice, legal documents). |
| RLHF — Reinforcement Learning from Human Feedback | التعلم التعزيزي من ردود الفعل البشرية | A training method where human evaluators rate the AI's responses, and the model learns to produce outputs that humans prefer. Used by most modern LLMs to improve safety and helpfulness. |
| Parameters | المعاملات / البارامترز | The numerical values inside a model that are adjusted during training. More parameters generally means a more capable model. GPT-4 has hundreds of billions of parameters. |
AI Agents & Systems
| Term (EN) | Arabic | Definition |
|---|
| AI Agent | وكيل الذكاء الاصطناعي | An AI system that can autonomously plan, make decisions, and take actions to complete goals — often using tools like web search, code execution, or file management. |
| MCP — Model Context Protocol | بروتوكول سياق النموذج | An open standard by Anthropic that allows AI models to securely connect with external tools and data sources, like Gmail, Slack, or databases — acting like a universal plugin system for AI. |
| API — Application Programming Interface | واجهة برمجة التطبيقات | A set of rules that allows different software applications to communicate with each other. AI APIs let developers build products on top of LLMs like Claude or GPT. |
| CLI — Command Line Interface | واجهة سطر الأوامر | A text-based interface where users type commands to interact with a computer or software — as opposed to clicking in a visual interface. Many AI developer tools are CLI-based. |
| Autonomous / Agentic AI | الذكاء الاصطناعي المستقل / الوكيل | AI systems that can execute multi-step tasks without human intervention at each step — browsing the web, writing files, sending emails, or running code on your behalf. |
| System Prompt | المطالبة الجذرية / السياق الأساسي | Instructions given to an AI model before the conversation begins, usually by the developer or platform, to define its role, tone, and behavior. |
Advanced Topics
| Term (EN) | Arabic | Definition |
|---|
| RAG — Retrieval-Augmented Generation | التوليد المعزَّز بالاسترجاع | A technique where the AI retrieves relevant information from an external knowledge base before generating a response, making answers more accurate and up-to-date. |
| Embedding | التضمين | A way of converting text (or other data) into a list of numbers (a vector) that captures its meaning. Embeddings let AI systems compare and find similar pieces of content. |
| Vector Database | قاعدة البيانات المتجهية | A database that stores embeddings (vectors) and allows AI systems to quickly search for semantically similar content — essential for RAG systems. |
| Inference | الاستدلال / التخمين | The process of using a trained AI model to generate outputs from new inputs. When you chat with Claude or ChatGPT, the model is running inference in real time. |
| Multimodal AI | الذكاء الاصطناعي متعدد الوسائط | AI models that can process and generate multiple types of content — text, images, audio, and video — in a single model. Examples: GPT-4V, Claude 3, Gemini Ultra. |
| Grounding | التأصيل / ربط النموذج بالواقع | Connecting an AI model to real-world, verified, and up-to-date information — through search, databases, or documents — to reduce hallucinations. |
| AI Safety | سلامة الذكاء الاصطناعي | The field focused on ensuring AI systems behave as intended, avoid causing harm, and remain aligned with human values — especially as models become more powerful. |
| Alignment | التوافق (مع القيم البشرية) | The challenge of ensuring that AI systems pursue goals that are beneficial and consistent with human values and intentions, not just optimized for narrow metrics. |
| Open Source AI | الذكاء الاصطناعي مفتوح المصدر | AI models whose code and/or weights are made publicly available, allowing anyone to run, modify, or build upon them. Examples: LLaMA, Mistral, Falcon. |
| Copilot | المساعد الذكي المدمج / كوبايلوت | An AI assistant embedded into a software tool (like Microsoft Copilot in Word/Excel, or GitHub Copilot in code editors) to help users work faster within that tool. |
Compiled by Conneqt Agency — Saudi Arabia's #1 AI MarTech Agency. Visit conneqtme.com for AI implementation services.