As a student myself, I am constantly exposed to the study techniques, habits and practices of my fellow learners. I see many different systems, and they would all be improved by incorporating AI to streamline the learning and writing experience. The purpose of AI in a student’s workflow is primarily to automate the tasks that don’t relate to directly learning content or producing written work.
This article covers some of these ways to incorporate AI into your education workflow. Examples of this include producing the flashcards that facilitate learning and providing feedback on an essay that can be acted upon to improve the final work. I present each use case in order of how they are used in my own work, so you can follow along to my workflow easier, and decide how these tools might help augment your own learning system.
Creating lecture notes
Lectures are the first point of contact for the content that most people learn at university. I turn up and am talked at for 45–50 minutes, and am expected to synthesise the information presented. I use AI to summarise the notes that I take in the lecture, so that when I’m revising, the content that I’m reading is as accurate and concise as possible.
I do this in Obsidian, using the QuickAdd plugin (detailed in this article), but there are other online tools that perform the same function, such as TLDR This and others. I prefer using QuickAdd in Obsidian, or ChatGPT or a LLM alternative, because of how you can manipulate the prompt to customise the summary generated.
In lectures, I currently use a transcription tool to record what the lecturer is saying, followed by a prompt that generates a summary of the important points covered. This way I don’t miss anything when I’m taking notes, and I have a comprehensive and accurate summary that I can feed into other steps of the process in the future.
Here is the prompt that I use to convert the transcription into a useful summary:
Please summarize and outline the following text.
It is an excerpt from a transcription, so cut out the speech-like language and make it a lot more concise.
Don't add any other messages aside from the outline to your response. Use only the text itself as material for summarization, and do not add anything new. Rewrite this for brevity, in an outline form.
Using bullet points, perform the above on the following text below this divider:
---
{{value}}
The `{{value}}` is a variable for the QuickAdd plugin in Obsidian, but you insert the transcription / lecture notes in place of this, so that the prompt contains the text that needs to be summarised.
The next sections talk about what we’re going to do with these summaries — we can automate further because the content isn’t yet flashcards which can be directly revised.
Creating flashcards
In my zoology course, much of the grade comes from examination of how well I know the course content. This means that I use active recall and spaced repetition techniques to learn the content the most efficiently. The means by which I create the flashcards is inconsequential in comparison to the effort I put in to learn them once created. This means that the flashcard production process can be automated using AI.
Again, I use a prompt in QuickAdd in Obsidian, but you can prompt ChatGPT to perform the exact same function for you. There is also an online tool called Paperclips that performs the same function as this prompt. Again, I prefer customising my own prompts as opposed to the web tool, because I use an Obsidian plugin to export to Anki for revision, and this requires flashcards to be formatted in a specific way. Here is the prompt that I use for flashcard generation:
Please create flashcards from the text below the divider ( - -) in the form:
Question/prompt here? #card
Answer here (don't add a full stop at the end of the sentence)
Leave a line break before adding the next card.
The question/prompt should not be double-barreled. Avoid general questions/prompts and double negative questions/prompts. Make sure that the question/prompt is unambiguous and relates precisely to the answer that I have to give. The answer should also be brief and simple, and entirely relate to the inputted content.
If you cannot contain the material in a single question/prompt, please create multiple flashcards.
Make the flashcards atomic, with one card pertaining to one 'atomic' piece of information. This is so I can better determine what parts of the content I need to revise or not. Create more cards than there are paragraphs if necessary.
---
{{value}}
This time, insert the summaries from the lectures and other content that you made using the steps and prompts from the previous section of the article.
This prompt provides a document of flashcards which I can then export to Anki for my revision. This workflow is now at the point that cannot be automated. You have to put in the time to learn the content of the cards that you’ve made. The guesswork has been greatly reduced by using a software like Anki though. You can forecast the number of new cards you need to learn per day leading up to exams, and the spaced repetition algorithm is based on how well you remember the content in reviews, so it makes it sure your time revising is being used as efficiently as possible.
Other AI uses for students
The above system greatly reduces the amount of effort required to learn content to a high level of understanding, but this isn’t the only job that you have as a student. You have to learn to write essays and longer pieces, and often you have to learn digital skills — for my course this is data analysis in R. AI can assist with these other jobs as well, and for the rest of this article I’m going to explain the other ways in which I automate and improve my education workflow.
I use ChatGPT to discuss topics that I don’t understand and can’t recover information about from my lecture notes — the prompt ‘explain as though talking to a child’ is useful in this situation. It’s also useful for debugging code that you may have to write as part of your course. I use it to check and format code for graphs I create for my lab reports.
There is some debate around using AI for writing in an educational context. Whilst I don’t think that you should submit content that is explicitly generated by AI tools, using AI to help with suggestions and improvements outside of the script you have written is a good idea. QuickAdd in Obsidian generates feedback on my writing — alternative viewpoints, changes and prompt questions, so I don’t ever run out of ideas. This helps me iterate my earlier drafts and improve the content that I submit.
Here is the prompt that I use to provide feedback on my academic writing:
I'm writing about {{value:Topic}}. Here's my draft:
{{value}}
Please suggest 3 improvements, 3 alternative viewpoints, and 3 questions for further reflection.
Again, replace `{{value:Topic}}` with your topic of interest, and `{{value}}` with the content of your piece.
Things to bear in mind
AI is useful for automating mundane tasks, but there is no shortcut to action towards an output or towards learning. This is in the sense that there is no suitable AI method for creating a good essay or lab report, and there’s certainly no way to automate the learning of course material. AI lets you work smarter by allowing you to apply attention in places where you’re going to have the most impact. It doesn’t completely negate the need to do work. Also be careful of sharing sensitive information with AI services.
Thank you for reading, please leave a clap and a comment if you enjoyed, and follow if you want to read more content like this.