AI terms in 2025 we help you to understand in simple plain language
AI terms in 2025 we help you to understand in simple plain language
- Generative AI: AI that creates text, images, etc.
- Chatbot: AI that simulates human conversation
- Compute: Processing power for AI models
- Computer Vision: AI that understands images and videos
- AGI (Artificial General Intelligence): AI that can do anything a human can do. AI that can think like humans.
- GPU: Hardware for fast AI processing
- Ground Truth: Verified data AI learns from
- AI Model: A trained system for a task
- AI Agents: Autonomous programs that make decisions.
- AI Wrapper: Simplifies interaction with AI models.
- AI Alignment: Ensuring AI follows human values
- CoT (Chain of Thought): AI thinking step-by-step.
- Fine-tuning: Improving AI with specific training data
- Hallucination: When AI generates false information.
- Context: Information AI retains for better responses.
- Context window: in AI is like the AI’s short-term memory
- Deep Learning: AI learning through layered neural networks.
- Embedding: Numeric representation of words for AI.
- Explainability: How AI decisions are understood.
- Foundation Model: Large AI model adaptable to tasks.
- Inference: AI making predictions on new data.
- LLM (Large Language Model): AI trained on vast text data.
- Machine Learning: AI improving from data experience.
- MCP (Model Context Protocol): Standard for external data access. Think USB-C for system to system
- NLP (Natural Language Processing): AI understanding human language.
- NLQ (Natural Language Query): free form text search box
- NLG (Natural Language Generation): AI generated language from inputs
- NLU (Natural Language Understanding): AI that helps machines understand what people mean when they talk or write.
- Guardrails: mechanisms put in place to ensure the safe and responsible use of LLMs
- Neural Network: A model inspired by the brain.
- Parameters: AI’s internal variables for learning.
- Prompt Engineering: Crafting inputs to guide AI output.
- Single shot prompt: clear and complete instruction in one go
- Multi-shot prompting: several examples or steps
- Reasoning Model: AI that follows logical thinking.
- Reinforcement Learning: AI learning from rewards and penalties.
- RAG (Retrieval-Augmented Generation): AI combining search with responses.
- Expert Augmented Generation: means giving AI a boost by adding expert knowledge
- Supervised Learning: AI trained on labelled data.
- Tokenization: Breaking text into smaller bits.
- Training: Teaching an AI by adjusting its parameters.
- Transformer: AI architecture for language processing.
- Unsupervised Learning: AI finding patterns in unlabeled data.
- Vibe Coding: AI-assisted coding via natural language prompts.
- Weights: Values that shape AI learning.