AI definitions

Basic AI concepts

Artificial Intelligence (AI)

The simulation of human intelligence processes by machines, especially computer systems. It encompasses various subfields, including machine learning, natural language processing, computer vision, and robotics.

Artificial General Intelligence (AGI)

A theoretical form of AI that possesses the ability to understand, learn, and apply intelligence across a wide range of tasks, similar to human cognitive abilities. AGI would be capable of performing any intellectual task that a human can do.

Machine Learning (ML)

A subset of AI that enables systems to learn from data and improve their performance over time without being explicitly programmed. It involves algorithms that can identify patterns and make decisions based on data.

Deep Learning

A specialized area of machine learning that uses neural networks with many layers (deep neural networks) to analyze data. It is particularly effective in tasks like image and speech recognition.

Neural Networks

Computational models inspired by the human brain, consisting of interconnected nodes (neurons) that process data. They are the foundation of deep learning.

Transformers

A type of model architecture that has revolutionized natural language processing and other fields. Transformers use self-attention mechanisms to process input data in parallel, enabling efficient training and high performance on tasks like translation and text generation.

Natural Language Processing (NLP)

A field of AI that focuses on the interaction between computers and human language. It enables machines to understand, interpret, generate, and respond to human language.

Large Language Models (LLMs)

Advanced NLP models trained on vast amounts of text data to generate human-like text. They can perform various language tasks, such as translation, summarization, question answering, and conversation.

Generative Pre-trained Transformer (GPT)

A specific type of large language model (LLM) developed by OpenAI that uses the transformer architecture. GPT models are pre-trained on diverse internet text and fine-tuned for tasks like dialogue, summarization, and code generation.

Computer Vision (CV)

A subfield of AI focused on enabling machines to interpret and understand visual information from the world, such as images and videos.

Reinforcement Learning (RL)

A type of machine learning where an agent learns by interacting with an environment and receiving rewards or penalties, improving its strategy over time.

Algorithms

Step-by-step procedures or formulas for solving problems. In AI, algorithms process data, learn from it, and make predictions or decisions.

Check some Machine Learning algorithms here

Retrieval-augmented generation (RAG)

RAG is a technique that combines a language model with a retrieval system. Instead of relying only on the model’s pre-trained knowledge, RAG allows the AI to fetch relevant external information (from documents, databases, etc.) in real-time and use it to generate more accurate, up-to-date, and context-aware responses.

Data

The information used to train machine learning models. The quality, quantity, and relevance of data significantly impact AI performance.

Model

It learns patterns from lots of examples, then uses those patterns to guess, decide, or create new things on its own. It is best thought of as a core element. It’s the central component that actually does the “thinking”: processing data, recognizing patterns, and producing outputs.

Difussion model

It is a generative approach that learns to create data (like images or audio) by reversing a gradual noise-adding process. It starts with random noise and iteratively denoises it, guided by learned probabilities, to produce realistic samples that match the training distribution.

Tensors

Mathematical objects used in deep learning to represent data. Tensors are multi-dimensional arrays that can hold various types of data, such as images, text, or numerical values.

Tokens

The basic units of text processed by NLP models. Tokens can be words, characters, or subwords, depending on the tokenization method used.

Training

The process of teaching a machine learning model using data. During training, the model learns to recognize patterns and make predictions based on input data.

Fine tuning

It is the process of adapting a pre-trained model to a specific task or dataset. Instead of training from scratch, the model’s existing knowledge is refined with new data, improving performance on specialized tasks while saving time, computation, and resources compared to full training.

It updates many (or all) of a model’s parameters, which can be computationally heavy and require large resources.

Checkpoint

It is essentially a saved snapshot of a trained neural network model at a specific point in time. When training AI models (like image generators, language models, or classifiers), the model gradually learns patterns by updating its internal parameters (weights). A checkpoint saves all those parameters into a full AI model.

LoRA (Low-Rank Adaptation)

It is a technique in AI for fine-tuning large models efficiently. Instead of updating all model parameters, LoRA inserts small trainable matrices into specific layers. This reduces memory and computation costs while preserving model quality, making customization faster, cheaper, and scalable for different tasks.

It only trains small, low-rank adapter layers added to the model while keeping the original weights frozen, making it far more efficient in memory, storage, and speed.

Inference

The phase where a trained model is used to make predictions or decisions based on new, unseen data, the application of the model in real-world scenarios.

Temperature

A parameter used in AI models, especially for text generation, that controls the randomness of predictions. A higher temperature leads to more varied outputs; a lower temperature makes outputs more deterministic.

Model Context Protocol (MCP)

It is an open-source standard designed to unify how AI models connect to external data and tools.