AI Terms Decoded: A Clear, Practical Glossary
TechCrunch offers a compact, approachable glossary of AI terms designed to cut through jargon and make key concepts accessible to everyone. Whether you’re a product manager, developer, or curious reader, the guide translates technical phrases into plain language so you can follow news, evaluate products, and participate in conversations about AI with confidence.
The glossary covers foundational concepts and hot-button topics alike. It defines what large language models (LLMs) are, explains phenomena such as hallucinations, and clarifies training techniques like reinforcement learning from human feedback (RLHF). The guide also highlights why these terms matter in real-world contexts—impact on reliability, user trust, and design choices for developers.
Here are a few sample entries from the glossary:
- LLM: A large language model trained on vast text datasets to generate and understand natural language.
- Hallucination: When a model produces incorrect or fabricated information despite sounding confident.
- RLHF: A method where human feedback is used to fine-tune a model’s behavior toward preferred outputs.
- Prompt: The input or instruction given to a model that shapes its response.
By translating jargon into clear explanations and practical examples, the glossary helps readers make smarter choices about using and evaluating AI. It’s a useful, positive resource for bringing more people into the conversation around AI’s development and deployment.