Using AI Tools Wisely
Get better answers from chat assistants, know when not to trust them, and protect your privacy while using them. Practical habits for study and work, with no coding.
Start this course →Lessons
- 1What a chat assistant really isKnow what is happening when you talk to an AI assistant.6 min
- 2Asking wellWrite prompts that get useful answers the first time.7 min
- 3Checking the answerKnow when and how to verify what an AI tells you.7 min
- 4Privacy and honest useUse AI tools without risking your data, your integrity or other people.6 min
What you will learn
- A chat assistant generates likely text one token at a time.
- It only sees what fits in its context window.
- Its built-in knowledge stops at a knowledge cutoff, unless it can search.
- A strong prompt states the goal, the context, the audience and the format.
- Few-shot prompting shows examples of what you want.
- Treat it as a conversation: refine, correct and ask follow-ups.
- Assistants can hallucinate: produce confident, false statements.
- For anything that matters, use verification against a primary source.
- Be most careful with numbers, quotes, references, recent events, and health, legal or money questions.
- Do not paste personal data, passwords or confidential material into AI tools unless you know how it is handled.
- Follow your school's or employer's rules on AI use, and be open about using it.
- Be sceptical of surprising images, videos and voices: they may be deepfakes.
Terms explained in this course
- chat assistant
- A program you talk to in everyday language, built on a large language model.
- prompt
- The message or instructions you give an AI tool.
- token
- A small piece of text, often part of a word, that a language model reads and writes one at a time.
- context window
- The amount of text a language model can take into account at once, measured in tokens. Anything beyond it is not seen.
- knowledge cutoff
- The date after which a model's training data stops. It may not know about later events unless it can search.
- few-shot prompting
- Showing a model a few examples of what you want inside the prompt, so it can copy the pattern.
- verification
- Checking a claim against a reliable, independent source before relying on it.
- primary source
- The original source of a fact, such as the official document, the research paper or the law itself.
- personal data
- Information about an identifiable person, such as a name, phone number, address or ID number.
- deepfake
- An image, video or voice recording made or changed by AI to show something that did not happen.