ChatGPT finally released their new feature, the code interpreter last Wednesday. We're super excited about the tool as it is a game-changer. It gives you the ability to upload documents and refer to them throughout your conversation, all without worrying about using up tokens.

But wait, there's more! We're not just here to tell you just what the code interpreter does. Instead, we're going to delve into the unconventional, the innovative, and the downright clever ways you can use this tool.

P.S. It's crazy.

What is the Code Interpreter Feature?

First things first, let's take a moment to understand what the code interpreter feature really is. The code interpreter is an innovative feature designed to make your AI interactions smoother and more versatile.

The magic lies in its ability to consume documents - yes, you can upload a whole document into ChatGPT! Once uploaded, you can refer back to this document throughout your conversation. The best part? It doesn't eat up your precious tokens. That means more room for meaningful and extended interactions.

But how does it work? The code interpreter uses a built-in interpreter to parse the document content. The content is then available to you in the conversation, just like another participant. You can ask questions, seek clarifications, and even dive into detailed discussions about the content. Sounds exciting enough right? Lets first look at it as what it was designed for, conventionally.

To unlock it, go to settings

To unlock it, go to settings in the bottom left corner.

And in Beta features, turn on the toggle for Code Interpreter.

Conventional Uses of Code Interpreter

Now that we've unlocked it, let's take a look at how the code interpreter is typically used. The conventional uses of this tool revolve around its primary function: executing code written in a high-level programming language. High-level languages are designed to be more understandable and easier to write for humans, but computers can't understand them directly. They understand only low-level code, like machine code or assembly code.

Here are some specific purposes of a code interpreter:

  1. Educational Purposes: It allows users to learn programming by directly executing code and seeing the results in real time. Users can experiment with different code snippets, learn from their mistakes, and understand the effects of various programming constructs.
  2. Troubleshooting Code: Users can ask ChatGPT for help with their code. If the code isn't working as expected, they can run it and ChatGPT can help identify the problem, suggest possible solutions, and then run the corrected code to ensure the problem is solved.
  3. Data Analysis: ChatGPT's code interpreter environment includes a number of popular Python libraries for data analysis and visualization, such as NumPy, pandas, and matplotlib. Users can load datasets, perform analyses, create visualizations, and more, all within the conversation.
  4. Demonstrating Concepts: When explaining a programming or data science concept, it's often helpful to provide a code example. With an integrated code interpreter, ChatGPT can not only provide the example but also run it, showing the user the output and helping to clarify the concept.

In summary, the code interpreter allows ChatGPT to be much more than just a text-based conversational agent. It can also act as a real-time, interactive coding assistant.

Unconventional Ways to Use Code Interpreter

When it comes to using ChatGPT, or AI in general, thinking out of the box is important. What are different ways you can connect words in the form of prompts to translate intent into output. While ChatGPT's Code Interpreter is great for code related tasks - at 4Fsh we have been using it far differently.

A Game Rulebook

One game that's pretty popular is Dungeons and Dragons. But with Core Rulebook being 320 pages long, that's 80,000 words. So we uploaded the rulebook to ChatGPT's Code Interpreter.

The output? We can ask ChatGPT everything within the rulebook, and have a conversation with the book instead of needing to flip through the pages to find what we need.

Resume Help

How about asking it directly about your resume? If you read our article about using ChatGPT as a mentor you would have seen how we used ChatGPT for an Artificial Intelligence Simulation based off of training data. The idea is that for more public figures, or at least people who have written a lot of content online, their data has been training LLM models.

So we uploaded a resume to the code interpreter and asked a digital Mike Podesto, a career expert to take a look at it. But why stop there? We can go further and ask make a dream team of multiple career experts and have them poke holes at a resume to improve it.

Remember, these are just two examples of unconventional ways to use the code interpreter. You can take it much further!

We're excited about ways that we can use the code interpreter tool. As it's currently in beta, it's bound to improve - but with what is right now available it should definitely be in your toolbox.

Have fun and Happy Prompting!