AI Glossary

AI systems that can autonomously plan, make decisions, and take actions to accomplish goals with minimal human intervention—the next evolution beyond chatbots.

A set of rules and protocols that allows software applications to communicate with AI services, enabling developers to integrate AI capabilities into their products.

The maximum amount of text (measured in tokens) that an AI model can process at once, functioning like the AI's 'working memory' for understanding and generating responses.

Numerical representations of text, images, or other data as vectors that capture semantic meaning, enabling AI to understand similarity and relationships between concepts.

The process of further training a pre-trained AI model on specific data to customize its behavior, knowledge, or style for particular use cases.

A large AI model trained on broad data that can be adapted to many tasks—the base layer upon which specialized AI applications are built.

When an AI generates confident but factually incorrect, made-up, or nonsensical information that isn't grounded in its training data or reality.

The process of using a trained AI model to generate predictions or outputs from new input data—essentially when the AI 'thinks' and produces a response.

The date until which an AI model has training data—any events or information after this date are unknown to the model without real-time retrieval.

The time delay between sending a request to an AI system and receiving the response—critical for real-time applications and user experience.

An AI system trained on massive amounts of text data that can understand, generate, and manipulate human language for tasks like writing, coding, and conversation.

AI systems that can understand and generate multiple types of data—text, images, audio, video—within a single model, enabling richer interactions.

AI models and tools whose code, weights, and training data are publicly available, allowing anyone to use, modify, and distribute them freely.

The practice of crafting effective instructions and context for AI models to get better, more accurate, and more useful outputs.

A technique that enhances AI responses by retrieving relevant information from external documents before generating an answer, reducing hallucinations and enabling real-time knowledge access.

A training technique that uses human preferences to fine-tune AI models, teaching them to generate responses that humans find helpful, harmless, and honest.

Compact AI models with fewer parameters that run efficiently on edge devices, offering faster responses and better privacy while sacrificing some capability compared to large models.

A parameter that controls the randomness and creativity of AI outputs—lower values produce more predictable responses, while higher values increase creativity and variation.

The basic units of text that AI models process—typically word pieces or characters. Token count determines pricing, speed, and context limits in AI tools.

The neural network architecture that powers modern AI language models, using 'attention' mechanisms to understand relationships between all words in a text simultaneously.

A development approach where programmers describe what they want in natural language, and AI generates the code—focusing on intent over syntax.