superuser_

Keywords bank

Every term the course defines, in plain language. Grows as more lessons land.

Bias
AI output reflects the average of its training data, including its skews. Average is not the same as good, fair, or true.
Confidence ≠ correctness
The key safety rule: how sure an AI sounds tells you nothing about whether it's right. Fluency is not evidence.
Context
The background you give the AI about your situation — who it's for, what you have, the constraints. It can't see any of this unless you say it.
Deterministic
Always giving the exact same output for the same input (like a calculator). An LLM is NOT deterministic — ask the same thing twice and the wording differs.
Few-shot (example-giving)
Showing the AI one or two examples of what you want. One example is worth a paragraph of description, because pattern-matching is what it does best.
Hallucination
When an AI gives a confident answer that is plausible-sounding but actually wrong, because it generates likely text rather than retrieving verified facts.
taught in: What an AI model actually is · L0see also: Next-word prediction
Iteration
Refining in the same conversation ('warmer, less jargon') instead of restarting. The conversation itself is the tool.
Large language model (LLM)
The technology behind tools like ChatGPT and Claude. A system trained on a huge amount of text that works by predicting the most likely next word.
taught in: What an AI model actually is · L0see also: Next-word prediction, Training data
Next-word prediction
The core mechanism of an LLM: given some text, it produces the most plausible continuation, one piece at a time. Reasoning-like behaviour emerges from doing this extremely well.
taught in: What an AI model actually is · L0see also: Large language model (LLM)
Prompt
What you send the AI. Not a search query — think of it as briefing a collaborator. Its quality sets the ceiling on the output.
Role prompting
Telling the AI who to be ('you are a patient tax accountant…'). It steers the answer toward the relevant region of what the model learned.
Training cutoff
The date an AI's knowledge stops. Plain chat has no live internet, clock, or calculator — so 'recent' or 'today' questions are unreliable without tools.
Training data
The body of text an AI learned from, up to a fixed 'cutoff' date. Out of the box it knows nothing after that date unless a tool lets it search.
Verify
Independently checking an AI's factual claims (and that its sources exist) before relying on them — a routine habit, not an afterthought.
taught in: What AI can't (and shouldn't) do · L0see also: Confidence ≠ correctness