Book a call

Daniel & Friends / Glossary / Hallucination

AI and automation

Hallucination

A hallucination is a case where a model states plausible-sounding but false information. That is why numbers, quotes and sources need checking.

Daniel Votruba ·


In practice

Add a checking step to the process: verify numbers at the source and trace quotes. A model that cites can just as easily cite something invented.

Related terms

  • Large language model (LLM): A large language model is a model trained on a huge amount of text that predicts the next words and can write, summarise and answer.
  • RAG (retrieval-augmented generation): RAG is a method in which a model first retrieves relevant documents, for example company knowledge, and answers from them.
  • AI slop: AI slop is low-quality content made by artificial intelligence in large quantities and with minimal human work.

More terms in the group AI and automation

Human in the loop · Guardrails · Inference and latency · Grounding · Jailbreak (AI) · Generative AI · Function calling (tool use) · Machine learning (ML) · Foundation model · MCP (Model Context Protocol) · Fine-tuning · Model bias (AI bias) · Few-shot and zero-shot prompting · Model evaluation (evals) · Explainable AI (XAI) · Multi-agent system · EU AI Act · Multimodal model · Embedding · Neural network · Diffusion models · NLP (natural language processing) · Deep learning · No-code and low-code

Full glossary (590 terms)

Terms in practice

Need help putting them to work?

Terms are one thing, putting them to work is another. We go through which of them pay off for you and say so plainly. You get a first opinion for free and we reply within 24 hours.

Book a call

Enjoyed the article? Share it with friends.