Aulafy01 / Foundations / Lesson 1 of 8
Understand what AI can do
Separate a useful draft from a fact you can trust.
01 / Understand
Start with a familiar task
Imagine asking someone to suggest a subject line for an email. You can judge the suggestion before sending it. Generative AI can help in a similar way: it produces text, images or other material from instructions. It is useful for making a first draft, reorganising information and exploring alternatives. It does not take responsibility for the result. You do.
A fluent answer can still be wrong
A language model generates text from patterns learned during training and the information supplied to it. It can produce an answer that sounds certain even when a name, number or explanation is wrong. This is often called a hallucination. Confidence in the wording is not a check of the facts. A claim needs evidence outside the claim itself.
Know the three parts
The model generates the response. The app is the interface where you type. Tools may let the app search the web or read a document. Different apps give a model different information and permissions. Do not assume an answer used current web information just because it contains a date. Look for evidence that a source was actually consulted.
02 / Worked example
Look closely at the result
A written teaching example, not a live AI response.
Suggest a short subject line for an email asking to move our team meeting.
Meeting moved to Friday at 10 am
Compare with the tutor’s review
The wording is short, but it invents a day and time and presents a request as a decision. A safer draft is: 'Request to reschedule our team meeting'.
03 / Your turn
Put it into practice
Write a safer subject line. Then explain what was invented in the example and who should approve the final meeting time.
Use fictional information. Nothing you write here is sent to a server or an AI. Download your work before leaving; it is not saved automatically.
04 / Check your understanding
One decision before you go
Take this with you
Use AI to create a candidate answer. Decide whether to trust it after checking.
Complete the practice, review each criterion and check your answer to finish your self-review. Every lesson remains open.
Next: Choose your first useful taskSources & course notes
Reviewed . Examples and exercises are original Aulafy teaching material under MIT. These references inform the course’s approach to reliability and learning; they are not endorsements.
- NIST: Generative AI risk profileBackground on unreliable outputs and evaluating risks.
- UNESCO: Generative AI in educationBackground on human agency, privacy and learning.