The AI Directory

🤯 Essential “Jargons” from A to Z

Algorithms: The core set of rules or instructions an AI model follows to process data, learn, and make predictions or decisions.

Artificial General Intelligence (AGI): A hypothetical form of AI with the capacity to understand, learn, and apply its intelligence to solve any problem, like a human being.

Autonomous Systems: Machines or software programs that can operate, make decisions, and achieve goals without continuous human control or input (e.g., self-driving cars).

Bias in AI: Systematic errors in an AI model’s output that arise from prejudiced assumptions or flawed data in its training set, leading to unfair results.

Big Data: Extremely large and complex datasets that traditional data processing applications cannot handle, which are essential for training modern AI models.

Chatbot: An AI program designed to simulate human conversation, typically through text or voice commands, to assist users with tasks or information retrieval.

Computer Vision: The field of AI that enables machines to “see,” interpret, and understand information from digital images, videos, and other visual inputs.

Deep Learning: A subset of Machine Learning that uses multi-layered Neural Networks (deep neural networks) to analyze complex data patterns, such as those in images or speech.

Ethical AI: The practice of designing, developing, and deploying AI systems with moral principles, ensuring fairness, transparency, accountability, and minimal harm.

Fine-Tuning: The process of taking a pre-trained large model (like an LLM) and training it further on a smaller, specific dataset to adapt it for a particular task or domain.

Generative AI (GenAI): AI models capable of creating new and original content, such as text, images, audio, or code, by learning the patterns from massive amounts of training data.

Hallucination: When an AI model, especially an LLM, generates information that is plausible-sounding but factually incorrect or completely nonsensical.

Large Language Model (LLM): A class of deep learning models trained on vast amounts of text data to understand, summarize, generate, and predict human-like language.

Machine Learning (ML): A core subfield of AI focused on building systems that learn directly from data and improve their performance on a task without being explicitly programmed.

Natural Language Processing (NLP): The branch of AI that gives machines the ability to read, understand, and derive meaning from human languages.

Neural Networks: Computing systems inspired by the structure and function of the human brain, featuring interconnected layers of “neurons” to process information and learn.

Prompt Engineering: The discipline of designing and refining the input (the “prompt”) given to a Generative AI model to achieve a desired, high-quality, and reliable output.

Reinforcement Learning (RL): A type of ML where an AI “agent” learns to make sequential decisions by interacting with an environment, receiving rewards for good actions and penalties for poor ones.

Robotics: The interdisciplinary branch of engineering and computer science that deals with the design, construction, operation, and application of robots, often powered by AI.

RAG (Retrieval-Augmented Generation): An advanced technique for LLMs where the model first retrieves relevant information from an external knowledge base before generating a final answer, improving accuracy and relevance.

Supervised Learning: A Machine Learning approach where the model is trained on a labeled dataset, meaning the input data is already paired with the correct output or “answer.”

Temperature: A hyperparameter in Generative AI that controls the randomness and creativity of the output, with higher values leading to more unpredictable results.

Token: The fundamental unit of text (which can be a word, part of a word, or punctuation mark) that LLMs use to process input and generate output.

Transformers: A neural network architecture, first introduced by Google in 2017, that forms the foundation for most modern LLMs and GenAI models due to its efficiency in processing sequential data.

Unsupervised Learning: A Machine Learning approach where the model is trained on unlabeled data, tasked with finding hidden patterns, clusters, or structures within the data on its own.