Your First Real Conversation

The honest case for caution.

What you'll leave with

By the end of this lesson, you'll understand AI's genuine limitations without becoming afraid of it — and you'll leave with calibrated trust, neither dismissive nor naive.

Why this matters

The previous lesson made a case for AI. This one balances it. A tool you do not understand is a tool you cannot trust — and trust built on false premises breaks at the worst moments.

The idea

AI gets things wrong in specific, predictable ways. Once you know the patterns, you can work with them.

The teaching block

It does not know what it does not know. This is the biggest one. AI can be confidently wrong. It does not say "I am not sure" reliably. It generates a plausible-sounding answer whether or not one exists.

Its training has a cutoff. AI was trained on data up to a certain point in time. Anything that happened after that date may be missing, outdated, or wrong. If you ask about current events, recent prices, new laws, or recent research, treat the answers with real caution.

It does not know your life. Unless you tell it your context, it is working from general patterns. It does not know your company, your relationships, your history, your constraints. Vague questions get generic answers.

It is not reliable for precise facts. Numbers, dates, citations, statistics — AI gets these wrong more often than it should. Never use an AI-generated statistic without checking it somewhere else.

It cannot replace your judgment. AI is good at language. It is not good at knowing what the right thing to do is in your specific situation. It can help you think. It should not decide.

It reflects its training data. That data has gaps, biases, and blind spots. It can reproduce assumptions and errors from the material it learned from.

Example

Ask AI a question about a recent event or a piece of very specific local knowledge. Show how it handles uncertainty — or does not.

Then contrast with asking it to draft an email, where it performs well. The contrast demonstrates the difference between language tasks and fact tasks.

Try this now

Ask AI something specific where you can easily check the answer — a recent news story, a local fact, something from your professional field. Note whether it is accurate. Note whether it acknowledged any uncertainty.

This is the beginning of the fact-checking habit that Module 5 covers in full.

Save this

Quiet takeaway

Knowing what AI gets wrong is useful. Understanding why it gets things wrong is what makes you genuinely skilled. That is the next lesson — and it is the most important idea in this module.

Next

Knowing what AI gets wrong is useful. Understanding why it gets things wrong is what makes you genuinely skilled. That is Lesson 5 — and it is the most important idea in this module.