The AI chatbot ChatGPT has become mega-popular in just weeks, faster than social media platforms like TikTok or Instagram. Only two months after its launch in late November, it had 100 million monthly active users in January. You probably heard that ChatGPT is at the moment the “must-have tool” for marketing teams or guest writers.

But how can ChatGPT speed up your daily work as a developer?

In this article, I describe some examples of ChatGPT usage from real programmers' cases. Usually, I'm programming in JS with TypeScript, so the examples will be on this stack. The theses presented in this article are subjective. Nevertheless, I would like to explore the possibility of using AI daily. 

So let’s start the exploring part

The first use case is the following: 

I would like to set up a new project for the AWS lambda function. It has to be written in TypeScript and based on the AWS CDK. Let’s see, what is a ChatGPT proposal: 

To create an AWS Lambda function in TypeScript based on the AWS CDK (Cloud Development Kit), you can follow these steps:

1. Install the AWS CDK and TypeScript:

2. Create a new AWS CDK project:

3. Install the AWS Lambda module:

4. In the lib directory, create a new TypeScript file for your Lambda function, e.g. my-lambda-function.ts.
5. Write your Lambda function code using TypeScript and the AWS CDK. Here is an example:

6. In the lambda directory, create a new TypeScript file for your Lambda function code, e.g. index.ts.
7. Write your Lambda function code using TypeScript. Here is an example:

8. Compile your TypeScript code to JavaScript:

9. Package your compiled JavaScript code and any other necessary files into a ZIP file:

10. Deploy your Lambda function using the AWS CDK:

That's it! You should now have a TypeScript-based AWS Lambda function up and running, deployed using the AWS CDK. Note that you may need to configure your Lambda function's runtime to use Node.js 14.x or later to support TypeScript.

The above content has been generated from ChatGTP. We have a complete guideline step by step how to create/set up the project with AWS lambda based on the CDK. This is absolutely awesome! Obviously, it’s probably not a super optimizable setup but usually, it depends on use cases. 

However, for me, it could be a good starting point extended in the future. The effort is absolutely minimum. 

ChatGPT can also be used to list the most common solution for the special aspect

Let’s define the next problem:

I would like to get a list of the most popular IVR providers. The AI response is like the below:

There are several IVR (Interactive Voice Response) providers available, each with their own set of features and pricing plans. Here are some of the most popular IVR providers:

1. Twilio: Twilio is a cloud communications platform that provides a wide range of communication APIs, including IVR. Twilio's IVR features include customizable voice prompts, call routing, and integration with other Twilio services.
2. Plivo: Plivo is a cloud communication platform that provides APIs for messaging, voice, and video calls. Plivo's IVR features include call routing, voice prompts, and integration with other Plivo services.
3. Nexmo: Nexmo is a cloud communication platform that provides APIs for messaging, voice, and video calls. Nexmo's IVR features include voice prompts, call routing, and integration with other Nexmo services.
4. Amazon Connect: Amazon Connect is a cloud-based contact center solution that includes IVR capabilities. Amazon Connect's IVR features include customizable voice prompts, call routing, and integration with other AWS services.
5. Voximplant: Voximplant is a cloud communication platform that provides APIs for messaging, voice, and video calls. Voximplant's IVR features include customizable voice prompts, call routing, and integration with other Voximplant services.
6. RingCentral: RingCentral is a cloud communication platform that provides APIs for messaging, voice, and video calls. RingCentral's IVR features include customizable voice prompts, call routing, and integration with other RingCentral services.
7. 8x8: 8x8 is a cloud communication platform that provides APIs for messaging, voice, and video calls. 8x8's IVR features include customizable voice prompts, call routing, and integration with other 8x8 services.

These are just some of the popular IVR providers available. When choosing an IVR provider, it's important to consider factors such as pricing, features, and ease of integration with your existing communication systems.

IVR implementation - Chat GPT proposal

So we can reduce a lot of time and instead of researching the information by ourselves - just asked the ChatGPT. Is it not awesome? But can we try to check the proposal for the IVR implementation?

I would like to implement the IVR using Pivo in TypeScript. Here is a result of that question:

Sure! Here's an example of how to implement a basic IVR (Interactive Voice Response) system using Plivo in TypeScript:

1. Install the Plivo SDK:

2. Import the Plivo classes:

3. Initialize the Plivo client with your credentials:

4. Create a function to handle incoming calls:

5. Create a function to handle user input:

6. Set up your routes:

That's it! This is a basic example of how to implement an IVR system using Plivo in TypeScript. You can customize the messages and actions to suit your needs.

Again the generated content looks quite nice. It’s a completed tutorial step by step how to start the implementation of the IVR based on the specific library. Obviously, you still need to check the documentation of the Plivo SDK but it can be a good starting point again. 

Can we use the ChatGTP to write the unit test for our code?

Let’s check the example such as:

This function goes from the helpers and is using to calculate the nights between two days. The bot response looks like the below;

Sure, here are some unit tests for the calculateNightsBetween function:

Some of the use cases have been covered. Also, the detection of the LocalDate was correct. The examples of the function hadn’t had information on what kind of library has been used LocalDate, but the chatGTP deducts it perfectly. 

I had to have a more strict test, so asked the Bot to add some extra variants

The problem was the following: 

Please add variants when one of the variables is undefined, null or empty. 

And again the response was quite nice:

Sure, here are additional tests to cover scenarios where one of the variables is undefined, null, or empty:

‍‍

Summary

Based solely on my opinion, I am inclined to assert that AI can be used in a programmer's typical day-to-day work. The examples I presented only show a fraction of the spectrum that we can utilize based on artificial intelligence, while also saving a lot of time. We must always remember that our problems often concern very specific issues, and therefore, the generated solution may not meet our expectations. However, AI should become one of the basic tools for a programmer, such as IDE or Stack Overflow.

We must always remember that our problems often concern very specific issues, and therefore, the generated solution may not meet our expectations. However, AI should become one of the basic tools for a programmer, such as IDE or Stack Overflow.

‍

Daniel