A sample mission · about 5 minutes · no account needed

Is a convincing answer also well supported?

Try the ScuolaAI approach here. This fictional case does not call an AI or assign a grade. You can work on paper or use the tools you normally use to read and write.

1. Start with a claim

«Questo assistente riduce del 70% il tempo di lavoro per chiunque. Lo abbiamo provato con tre colleghi per una settimana.»

You are responsible for deciding whether to try it with your team. Which part of this claim does the information support? What check is missing?

Give me a hint

Separate what was observed from what is being generalized. Do three colleagues represent all people and all types of work? How was time measured?

Show me a similar example

“Two people understood the topic better after watching a video, so videos always work better than books.” Their results may be real, but they do not support “always.” A useful check compares similar activities, prior knowledge, and understanding after a delay. Now return to the assistant: what comparison would you propose?

2. Build and check

Rewrite the conclusion within the limits of the data. Propose a small test: who would you include, what would you measure, and what result would make you change your mind? A concise answer is fine.

Check your reasoning against the criteria

• Did you distinguish the observation from the generalization?
• Did you identify missing information without making it up?
• Did you propose a practical comparison and a result that could disprove the hypothesis?

The right words alone are not enough: look for evidence of each criterion in your answer. You can defend a different choice as long as you explain its limits.

3. Try a new case, without hints

“All students learn better with an AI Tutor: five volunteers finished the course ahead of schedule.” What is missing to support this conclusion? Identify evidence of learning that is different from speed of completion.

In a week, try again with a different news story. Can you still spot the leap from data to conclusion? In the full program, you keep your work, checks, declared assistance, and reflection. An assessment with reasons explains what the evidence supports and what still needs work.

Create an account Read the manifesto

Sign-in and the learning program are currently available in Italian. An English version is planned.