Module 2 of 4Lesson 1 of 2~10 min
What an AI assistant really does
An AI assistant can make things up, forget what you told it or miss recent news. This lesson explains in plain terms why these errors happen and helps you choose, for each one, the right way to protect yourself from it in your everyday work with these tools.
Lesson objective
By the end of this lesson, you will be able to explain why an AI assistant makes things up, forgets or misses recent information, and choose the right fix for each case.
Topics covered
- how LLMs work
- hallucinations
- AI assistant
- AI limitations
Where it fits
Understand the tool before you use it
Why does an AI assistant get things wrong, and what can I safely give it?
Lessons in this module
- What an AI assistant really does (this lesson)
- Decide which data to give to AI
What you will learn in the course
This lesson is part of the course Put AI to work in your business
- Explain, on a concrete case, why an AI assistant makes things up, forgets or misses recent information, and decide how to use it accordingly.
- Sort data before giving it to an AI tool (allowed, anonymize first, forbidden) and configure the tool to match.
- Choose which tasks to hand to AI and how to work with it (automate, assist, delegate) based on value, risk and the data involved.
- Write a complete request (task, context, examples, format, constraints) and improve it over at least two documented iterations.
- Pick the right tool for a deliverable and justify the choice on data, cost, licensing and integration.
- Check AI output against an explicit checklist (facts, figures, sources, tone, compliance) before using it.
- Turn a successful use into a reusable workflow with human checkpoints, and set the team's rules of use in an AI policy.
Related courses
- Understand what LLMs do well, and where they failAll levels · ~2 hr
- Choosing and using AI tools at workAll levels · ~1 hr 40 min
- Prompting for product managers: delegate, describe, verifyJunior · ~2 hr 45 min