Foundation · About 20 minutes with practice

Practice until the idea becomes yours

Use retrieval, feedback and transfer to turn a compelling explanation into something you can actually use.

Make this yours

A learning record that preserves your first answer, a correction and a new-context application.

Start with your own answer ↓
Speculative Academy artwork: the hall where scale becomes tangible.
Imagined world · The hall where scale becomes tangible

A moment before the explanation

What is the difference between recognizing an explanation and retrieving it?

Give yourself a first answer, even if it is incomplete. Keep it in mind as you read—or write it in your field note.

01

Try to retrieve before you reread

An explanation can feel familiar without being available when you need it. Close the source and state the idea in your own words. Then compare your answer with the reference. The gap tells you what to practice next more usefully than another effortless reread.

The Academy places a recall prompt before the editable artifact. You may be uncertain; write that uncertainty down. A wrong first attempt can still be valuable if the feedback is specific and you preserve the correction. The aim is to make thinking visible, not to reward speed.

02

Ask for feedback that can change a decision

“Excellent work” tells you very little. Ask a reviewer to point to a claim, missing step or counterexample. Preserve your original answer so you can tell whether the revision addresses that exact issue. A model can offer feedback, but its feedback should also be checked.

Separate observations about behavior from assumptions about a person. “The reviewer skipped two references” is inspectable; “the reviewer is careless” is a broader interpretation. This distinction supports humane collaboration and makes improvement more specific. Model fluency and human confidence both need calibration.

03

Change the context and test the idea again

After revising, apply the same principle to a different situation. A permission boundary learned in a research task should be rethought for a customer message. A graph designed for documents may need different relationships for organisms or artworks. Transfer reveals whether you learned a phrase or a usable distinction.

Return after a gap and try the recall prompt again. This site does not infer mastery from time spent or a saved checkbox. You choose when to save a field note and when you have reviewed it. Keep an unfinished question if that is the most honest result of the session.

A worked example · Illustrative

A useful revision record

First answer: A successful tool response means the task is done.
Feedback: Which evidence shows the result meets the goal?
Revision: The response proves only a tool-level event.
Transfer: A saved file can still contain unsupported claims.
Next retrieval: explain this distinction without the example.

Put the idea to work

Your practice

  1. Answer the recall prompt before reopening the explanation.
  2. Ask for one precise counterexample to your first answer.
  3. Save the original and the revision side by side.
  4. Apply the revised idea to another domain and note what changes.

Make the idea available when you need it

Your field note

Keep your first answer, make an artifact and try the idea somewhere new. Save only what you want stored on this device.

Check one distinction

You remember the phrase “human in the loop” but cannot specify who decides what. What would help most?

Take this practice into your own agent.

Copy a coaching brief, then choose your host in the Academy installer.

Choose a host ↗

Your text stays in this page until you save or export. Local saving can be unavailable in restricted browsers. Curriculum 2026-09-15.1. Import, export and browse your notebook ↗

Follow the evidence further

Primary sources & deeper study

Selected 15 September 2026. Provider docs govern current setup; older courses and engineering articles provide foundations.

RetrievalPractice.org · Research-led teaching resource

Retrieval practice

Practice bringing an idea to mind, then use feedback. Our exercises adapt this principle.

RetrievalPractice.org · Research-led teaching resource

Metacognition and feedback

Use feedback to compare confidence with evidence.

Harvard · CS50 · Open course

CS50 introduction to AI with Python

Hands-on foundations in search, knowledge, uncertainty and learning.

Take the method into your work

Practice a revision with feedback
Another question to follow

Give your first agent a real job