Programming with AI productivity boost #3: Kickstarting new projects

I love exploring new frameworks, standards, and programming languages. However, setting this up in a new project can be time-consuming. Recently, I started using AI to streamline the process. By crafting a prompt that describes my use case and desired technologies, followed by the phrase "give me the commands to get started", I can quickly obtain easy-to-follow instructions from ChatGPT to jumpstart my project.

Preparing for my upcoming session at Techorama in October, where I'll be discussing how you can use AI to build your frontend, I wanted to set up the backend using Microsoft technologies. So I described my project requirements to ChatGPT as follows:

I want to create an e-learning web application "LearnFromAI" using modern dotnet core 8. It uses a sql database, entityFramework and users need to login. Everything must be production-ready and run in Docker. I want to develop using a devcontainer in vscode and I want to use powershell.

While ChatGPT confidently provided a 10-step plan, it quickly became evident that ChatGPT had made some missteps. Upgrading a database schema before configuring the database for example is evidently wrong.

I do not give up easily, so I did two things get a better plan. First of all, I switched from ChatGPT 4o to Claude 3 Opus. Second, I used a better engineered prompt. So I asked the best prompt engineer I know, 🥁..., AI 😆 to write me a better prompt by giving the project explanation followed by "What would be the best prompt for chatgpt to get it to help me set this up one step at the time?". The resulting prompt was the following:

I want to create a production-ready e-learning web application called 'LearnFromAI' using the following technologies:

.NET Core 8

SQL database

Entity Framework

User login functionality

Docker for containerization

Visual Studio Code with a devcontainer

PowerShell for development

Please guide me through the process step by step. For each step, provide the necessary commands, code snippets, and explanations. Let's start with setting up the project structure and progress towards implementing the features. Please ensure that the application is production-ready and can be run using Docker.

Let's begin with the first step. What should I do to set up my development environment with Visual Studio Code, a devcontainer, and PowerShell?

Using the improved prompt, Opus confidently guided me through the necessary steps, providing commands, code snippets, and explanations along the way. You can find the whole conversation here.

Although there were a few errors along the way, Opus immediately knew the solutions, and we successfully completed the project setup. Without AI assistance, this process would have taken me significantly longer with more frustrations. Moreover, I now have confidence that the project adheres to commonly used conventions and naming practices.

The next time you need to start a project, create a proof of concept, or explore new technologies, remember that AI can boost your productivity by guiding you through the process. While the generated project definitely has some shortcomings, which I plan to address in a follow-up post, it serves as a solid foundation for further development.