How to make a study guide and practice quiz from a lecture

What a study pack contains

StreamSnag calls the whole thing a study pack. The guide opens with the big picture and what you should be able to do afterwards. Then come the key ideas, each as a question with an explanation and the lecturer's own words. After that: review sheets for the hardest topics, key figures from the slides, easily confused pairs, formulas and numbers, a self-test, a practice quiz with every wrong option explained, and a glossary. Flashcards come with it and have a guide of their own.

Every item names the minute and the slide it came from. Open a sample study guide to see all of it; AI wrote it from an invented lecture.

The steps

1. Save the lecture and get its transcript

   The pack is written from the transcript and the text on the slides, so it needs both steps before it on the lecture's path. Slides are optional but make the guide better.

2. Choose an AI account

   Open Settings. On the AI tab, pick the provider and sign in or paste a key. AI stays off until you do this.

3. Press Make study pack

   It is the fourth step on the lecture's path, which “Show the path” opens. Before anything is sent, a sheet headed “Before the AI reads this lecture” shows what it makes, what it sends, to which account, and roughly how much AI use it takes. You can leave out parts you do not want. The button at the bottom is “Make the study pack”; “Not now” sends nothing.

4. Let it run

   It takes one to three minutes and carries on if you close the popup. It reads the lecture, writes the guide, writes the cards and questions, then checks every card against the lecture and saves the result next to the video.

5. Open the guide and do the quiz

   Answer the self-test from memory first. The practice quiz gives new cases to try the ideas on, tells you why each wrong option is wrong, and keeps a score.

How to tell whether to trust it

AI can write notes that sound right and are not. The pack is built to limit that: it first pulls out facts together with the lecturer's exact words, writes from those, and a last pass removes statements the lecture does not support. In the statistics sample, 61 of 65 statements passed the quote check and the rest were dropped. Where the guide explains something from outside the lecture, it does so in a box labelled “Beyond the lecture: not from your lecturer”. Those boxes can be switched off in Settings.

That lowers the error rate. It does not make it zero. So use the minutes: when an explanation surprises you, click its minute and hear the lecturer say it. The guide also has a section called Check this, which lists terms the transcript may have misheard.

What can go wrong

• The AI allowance runs out. A free allowance can do part of a pack and stop. StreamSnag says why it stopped. Run it again later or choose another account.
• A poor transcript. The guide can only be as good as the words it was given. Misheard names carry through, which is another reason to give it the slides.
• Your course's rules. Some universities restrict putting lecture material into AI tools. StreamSnag tells you this once. Whether to do it is your decision.
• A long lecture costs more. The estimate before each run grows with the lecture. Check it before you press the button.

Questions