Remember when you were in school and would create flashcards for yourself?  Or maybe you even purchased some to help you study for an exam.  Or for a professional certification at your job.  Well, today, look no further than ChatGPT to generate flashcards and help you study.

Tell ChatGPT what you’re studying, and how you want the questions formatted.

And voila!, your first flashcard:

Questions too easy?  Too difficult?  You can adjust that too with prompts like, “Make the next question easier,” or, “Make future questions at an advanced level.”

NOTE: If you don’t already know, ChatGPT can sometimes hallucinate (i.e., give incorrect information).  So make sure you do a sanity check on any “correct” answers that seem a little bit off.

Perhaps the most obvious way to use ChatGPT is to have it explain and build on concepts.  Maybe you’re new to Flask, and need a beginner-level understanding.

Then you build on it, getting into more detail that’s relevant for a developer.

You can also use prompts like, “Explain it for a 5-year-old,” “Explain it for a teenager,” and so on.  Just ask for what you want, and at what level.

Whether you’re a full-time researcher or just need to occasionally research things for school or work, ChatGPT can help find reputable sources to get you started.

In your prompt, indicate the topic you’re researching, and then ask for sources.

ChatGPT will come back with a summary of the topic…

…along with sources you can use to dig deeper:

NOTE: As always, be sure to validate the sources.  You don’t want to cite sources in a research paper that don’t actually exist!  That would be embarrassing.

Text extraction and summarization are areas where ChatGPT really shines.  Whether you’re sifting through research papers, online articles, or internal documents at work, it would be nice to get a high-level summary without having to read every word.

Here are some prompts you can use to do that:

Sometimes when you’re trying to learn something new, you actually need to understand two or more things and how they relate to each other.  ChatGPT is really helpful when it comes to comparing options, detailing pros and cons, and giving examples.

If you already understand one of the options pretty well, you can ask for an explanation of the other as it relates to the first. Like, “I am familiar with AWS CloudFormation.  Can you explain Terraform for me as it relates to CloudFormation?”

By now, you surely know that ChatGPT can be used to write code.  But to make that code even more useful from a learning perspective, you can ask it to explain what the code is doing too.

And after ChatGPT outputs the code, it will include a separate section that explains what the code does:

Note: Sometimes, ChatGPT will automatically give you explanations for the code it writes, but other times, it will only give you the code.  So if you know that you want the explanations, just ask for them up front.

How often do you inherit someone else’s code, or find something on Stack Overflow and wonder what it’s doing?  (Or let’s be honest, how often do you wonder the same thing about code you wrote a few months ago?)

Building on the last section, you can also feed existing code into ChatGPT to figure out what the code is doing.

Once ChatGPT explains the code, you can have a back-and-forth conversation to clarify anything that’s still unclear, asking about specific libraries, versions, syntax and so on, until you have a good understanding.

Sometimes you learn best when it all goes wrong, and that’s certainly true with writing and debugging code.  In the past, you would likely spend hours searching on Google or Stack Overflow, stepping through a debugger, and doing lots of trial-and-error to figure out why something wasn’t working.  In the end, you likely learned something valuable.  With ChatGPT, you have a very intelligent pair programmer who can help you diagnose and fix bugs in much less time.

If you have a specific error message, you can pass that into ChatGPT and get suggestions on how to fix it.  And if you have a general idea of the code that’s causing the problem, it’s helpful to paste that in as well.

Depending on the error, ChatGPT will usually suggest possible reasons for the error.  If the solution is a change to your code, it will also give you the updated code.

If you don’t have a specific error message, you can just describe the problem you’re encountering, and ChatGPT will help you identify and fix the problem.

This works for all kinds of issues beyond just code—problems with permissions, ports being closed, missing libraries, integration between parts of an application and a lot more.  But regardless of the problem, the more detail you can provide, the better troubleshooting help you’ll get from ChatGPT.

We’ve seen how ChatGPT can help learn a new coding language, but it also excels in human language as well.  Taking a sabbatical to France next year, and need to learn some French?  Or do you just want to impress your teammates in different parts of the world with your conversational skills in their native language?

You can start by asking for a basic approach to learning the language:

Then as you’re learning words and phrases, ChatGPT can act as a tutor to help check your understanding: