Learn AI with questions

Pick an answer, then see why. In 12 small steps, find out how a model turns into a system that gets work done.

From answer to action1 / 12

STEP 1 · WE ARE BUILDING A MEETUP PAGE

You are building a sign-up page for a coffee meetup. You typed what you want into an AI assistant. In this example the model has no access to the files on your computer or to your website.

The model wrote the code for the page. Does that mean the site is live?

Pick an option. Right or wrong, we will go through the explanation together.

How does this quiz teach?

The questions build on each other. Every option has an explanation, so when you get one wrong you also see the right approach. An explanation opens after each answer, and you move on to the next question yourself. There is no automatic countdown.

Here, "harness" means the layer of execution and coordination around the model. The event-page example was written for teaching. The sources were reviewed on November 1, 2026.

Your answers are kept only on this open page and are cleared if you refresh. Your learning score is not saved to your account or to any contest leaderboard.

Question sources (5)

Building effective agents Models, tools, workflows and the agent loop.
Writing effective tools for agents Tool descriptions, inputs and meaningful results.
Effective context engineering for AI agents Selecting relevant information for the context and managing it.
Effective harnesses for long-running agents Working state, continuing across sessions, and testing.
Demystifying evals for AI agents What a harness is, verifying outcomes and repeated evaluation.

All contests