◉ ai

Hallucination

When AI confidently generates incorrect information

When an AI model generates confident but factually incorrect, fabricated, or nonsensical information. LLMs don't "know" facts — they predict likely text sequences, which means they can sound authoritative while being completely wrong. For content creators and consultants, hallucinations are the biggest risk when using AI for research or drafting — always verify claims, especially around token prices, protocol specs, and regulatory details.

Related terms

RAG (Retrieval-Augmented Generation)

Giving AI access to your own data

LLM (Large Language Model)

AI models that understand and generate text

Prompt Engineering

Crafting inputs to get the best AI outputs