Artificial intelligence (AI)
Sounds like: robots taking over
Software that does things we normally associate with human thinking, like understanding language, recognizing images or making predictions.
The Nonsense Dictionary
22 AI buzzwords translated into plain English. What they sound like, and what they actually mean.
Machine learning
Sounds like: a machine going to school
A way of building AI where the computer learns patterns from lots of examples instead of following rules a person wrote by hand.
Neural network
Sounds like: a digital brain
A kind of machine learning model made of many simple math units in layers. It is loosely inspired by the brain, but it is just math.
Generative AI
Sounds like: AI that generates hype
AI that creates new content, such as text, images, audio or code, based on patterns it learned from existing examples.
Large language model (LLM)
Sounds like: a very big dictionary
An AI model trained on huge amounts of text to predict the next word. ChatGPT, Claude and Gemini are built on LLMs.
Chatbot
Sounds like: a robot in a chat window
A program you talk to in plain language. Modern chatbots usually run on a large language model.
Token
Sounds like: crypto?
A chunk of a word. AI reads and writes text in little pieces, not whole words or single letters.
Context window
Sounds like: a browser thing
How much text the AI can keep in mind at once, including your messages and its replies. Its short-term memory.
Prompt
Sounds like: very technical
The thing you type to the AI. Yes, that is it. That is the whole thing.
Hallucination
Sounds like: a medical emergency
When the AI makes something up and says it with total confidence. It predicts plausible text, not checked facts.
Training data
Sounds like: a gym for robots
The examples an AI model learns from. Its quality shapes everything the model knows and gets wrong.
Parameters
Sounds like: settings you can change
The millions or billions of numbers inside a model that get adjusted during training. They store what the model learned.
Temperature
Sounds like: the AI is overheating
A setting that controls how random the AI's answers are. Low means safe and predictable, high means more creative and more chaotic.
Embedding
Sounds like: something in a news report
A way of turning words or documents into lists of numbers so a computer can measure how similar their meanings are.
RAG (retrieval-augmented generation)
Sounds like: a cleaning cloth
Letting the AI look things up in your documents or a database before it answers, so it relies less on memory.
Fine-tuning
Sounds like: a piano lesson
Extra training on a smaller, focused set of examples so a model gets really good at one specific job.
Inference
Sounds like: detective work
The moment the AI actually runs and produces an answer, as opposed to when it was being trained.
Multimodal
Sounds like: a transport app
An AI that can handle more than one kind of input or output, such as text, images, audio or video.
AI agent (agentic AI)
Sounds like: a spy
An AI that does not just answer but takes steps for you, like searching, clicking or using other tools, to finish a task.
Prompt injection
Sounds like: a vaccine for prompts
A trick where hidden instructions in a web page, email or file try to make an AI do something its user did not ask for.
Bias
Sounds like: an opinion
When an AI's output is unfair or skewed because of patterns in its training data or how it was built.
AGI (artificial general intelligence)
Sounds like: next Tuesday
A hypothetical AI that could learn and do any intellectual task a human can. It does not exist yet, and experts disagree on when or if it will.
Missing a buzzword? Ask us, there are no dumb questions.
Ask a Dumb Question