An input or question given to an AI system to generate a response or perform a task, guiding its behaviour to provide relevant and useful information.
Incorporating human oversight or intervention in an AI system's processes to ensure accuracy, make critical decisions, and address complex situations that the AI alone cannot handle.
Fine-tuning a large language model by adjusting its training on specific datasets to improve its performance and relevance for tasks or domains.
A type of machine learning where an AI system analyses unlabelled data to identify patterns, relationships, or structures without predefined outcomes.
A type of machine learning where an AI system is trained on labelled data, meaning the inputs are paired with correct outputs, allowing the system to learn and predict outcomes for new, similar data.
A technology that enables AI systems to identify and classify objects, patterns, or features within images, allowing for automated analysis and understanding of visuals content.
Built-in guidelines and constraints within AI systems designed to ensure safe, ethical, and reliable operation, preventing misuse and minimizing errors.
Complex and unexpected actions or patterns that arise from simple rules or interactions within an AI system, not explicitly programmed by developers.
A set of step-by-step instructions or rules designed to perform a task or solve a problem, guiding how an AI system processes data and makes decisions.
When an AI system generates incorrect or nonsensical information that appears plausible but is not based on the provided data or reality.
A branch of artificial intelligence that enables computers to understand, interpret, and respond to human language, enhancing communication between humans and machines.
An AI-powered software application designed to imitate human conversation, allowing businesses to automate customer interactions and provide support through messaging platforms.
Unstructured data is information that lacks a predefined format or organization, such as text, images, and videos, making it more challenging for computers to process and analyse directly.
Organized information, typically in tabular format, that is easily searchable and analysable by computers, such as databases or spreadsheets.
The process of crafting and refining input prompts to guide an AI system's responses effectively, ensuring accurate and relevant outputs for specific tasks or applications.
A type of artificial intelligence that enables systems to learn from data and improve their performance over time without being explicitly programmed for specific tasks.
A type of artificial intelligence that can understand and generate human language, helping businesses with tasks like customer service, content creation, and data analysis. LLMs are trained on huge sets of data. Examples of LLMs include GPT, Claude and Gemini.
Artificial intelligence systems that can create new content, such as text, images, or music, by learning patterns from existing data and generating original outputs based on that knowledge.
Large, pre-trained AI systems designed to understand and generate a wide range of data, providing a base for further fine-tuning to specific tasks or applications.
Subset of machine learning that uses deep neural networks to analyse and learn from large amounts of data, enabling complex pattern recognition and decision-making.
A set of protocols and tools that allows different software applications to communicate and interact with each other, enabling seamless integration and functionality.
A type of large language model that uses deep learning to understand and generate human-like text based on patterns learned from extensive datasets.
Set of algorithms that have been trained on data to recognize patterns and make predictions or decisions based on that data.
