AI lexicon
Artificial Intelligence Glossary
The essential AI terms, explained simply and without jargon. So you understand what everyone is talking about.
- Artificial intelligence (AI)
- Computer systems capable of carrying out tasks that normally require human intelligence: understanding text, generating images, making decisions.
- Language model (LLM)
- A model trained on huge amounts of text to understand and generate language. ChatGPT and its counterparts are examples.
- Prompt
- The instruction you give an AI. The quality of the prompt largely determines the quality of the answer.
- AI agent
- An AI that doesn't just answer but gets tasks done: it chains steps together and uses tools.
- Hallucination
- When an AI invents false information and presents it as true. Hence the importance of human verification.
- Machine learning
- Systems that learn from data rather than being programmed rule by rule.
- Fine-tuning
- Specializing an existing model on your own data to adapt it to your business.
- RAG
- Retrieval-Augmented Generation: connecting an AI to your documents so it answers from your real data.
- Open source
- Models whose code is open: you can host and run them yourself, the key to data sovereignty.
- Inference
- The moment the model processes your request and produces an answer. Where it happens (EU or US) matters for sovereignty.
- Token
- The unit of text a model processes (a piece of a word). Billing and limits are often expressed in tokens.
- API
- An interface that lets your software communicate with an AI in an automated way.
- Data sovereignty
- Keeping control over where your data is stored and processed, and over who can access it.
- Automation
- Handing repetitive tasks over to systems to free up human time. AI expands what can be automated.
- LPD / nLPD
- The Swiss data protection act, which governs the processing of personal data.
Ready to move to AI?
Build your website with AI, or explore AI in Switzerland.