Lesson 1 — What AI Assistants Actually Are
Start by doing the thing yourself — no explanation yet. Read each phrase and pick the word you think comes next.
This activity needs JavaScript enabled. The lesson below still covers everything.
What you just did
You looked at some text and picked the most plausible word to come next. That is the entire job of an AI assistant.
Claude, ChatGPT, Gemini, and Copilot are all large language models (LLMs). An LLM is a program trained on an enormous amount of text. From that text it learned the patterns of language well enough to continue any piece of writing in a way that is usually sensible, relevant, and correct.
That is the whole trick: given everything written so far, the model predicts what should come next. It is not looking facts up in a database and it is not reasoning the way a calculator does. It is producing the most plausible continuation. Most of the time the most plausible continuation is also the correct one — but not always, and that gap is the single most important idea in this lesson.
Good at it, or not? You decide
For each task below, choose whether an AI assistant is reliably good at it. You will get instant feedback — guessing is fine, that is how you learn the pattern.
This activity needs JavaScript enabled. The lesson below still covers everything.
Why the misses happen
The tasks an AI struggles with are not random — they share a cause. A model predicts plausible text; it does not check that text against reality. So it fails in predictable ways:
- Confident mistakes. A model can state something false in exactly the same calm, fluent tone it uses for the truth. There is no built-in "I am unsure" signal you can rely on.
- Made-up details. Functions, citations, or settings that sound real but do not exist. This is called a hallucination.
- Stale knowledge. Training data has a cutoff date, so a model may not know about recent changes — including recent changes to your own project.
- Exact arithmetic and counting. Improving, but still a weak spot. Verify numbers.
None of this makes the tools unreliable — it makes them tools. You get the speed; you supply the judgment.
The main assistants you will hear about
| Tool | Made by | Typical use |
|---|---|---|
| Claude | Anthropic | Long documents, careful reasoning, and coding. |
| ChatGPT | OpenAI | General-purpose assistant, widely used. |
| Gemini | General-purpose, integrated with Google products. | |
| Copilot | GitHub / Microsoft | Code completion inside an editor. |
They differ in detail, but the mental model in this lesson applies to all of them. Learn one well and you can move to any of them.
Quick check
This activity needs JavaScript enabled.