Guest Blogs

Blog archive

Partners: Get on the GitHub Copilot Wagon Now

I'm not a software developer; I'm just a guy helping Microsoft partners find their path to success. Nitty-gritty technical conversations have never been my arena.

But even a non-techie like myself sees that AI-driven software development is a revolution. Software development will never be the same again, to the benefit of both customers and partners. And this is where I'm an expert. I love to talk about business transformation and driving better business outcomes.

We can thank GitHub Copilot for this new era of software development. To say that GitHub Copilot can help you write code faster and better is an understatement of giant proportions. With GitHub Copilot, you can perhaps get 80 percent of your code generated and can spend time really focusing on the remaining 20 percent.

Here are some ways GitHub Copilot (or current and future competitive solutions) can reduce time and cost in software development:

  • It can generate code from natural language descriptions. You can use comments or prompts to describe what you want to do, and GitHub Copilot will suggest relevant code snippets or functions that match your intention. This can save you time searching for existing solutions or writing code from scratch.
  • It can learn from your code and style. GitHub Copilot can adapt to your project's context and style conventions, and offer suggestions that are consistent with your codebase. This can improve the quality and readability of your code, and reduce the need for refactoring or debugging.
  • It can help you discover new APIs and libraries. GitHub Copilot can suggest code that uses popular or relevant APIs and libraries for your project, and even provide documentation or examples for them. This can help you explore new possibilities and features for your project, and reduce the learning curve for new technologies.
  • It can help you test and optimize your code. GitHub Copilot can generate test cases, benchmarks or performance metrics for your code, and suggest ways to improve or optimize it. This can help you ensure the reliability and efficiency of your code, and reduce the risk of errors or bugs.

There might be people out there who say that traditional software development is far better than taking help from AI, or that using AI is like cheating.

My response would be that I've heard similar things before. The first time was when I was young and desktop publishing with the Apple Macintosh was just becoming popular, and the oldies said that it was not good enough. All of us also heard it when cloud computing started to become popular.

But guess what? As these new technologies became got more popular and refined, we haven't looked back.

If I were running a software development practice -- either as an ISV or as more of a custom development -- I would transform the way we work and use GitHub Copilot right away. In so doing, I could significantly increase my team's output and perhaps reduce the number of developers.

This opportunity is, of course, larger in developed economies with higher salaries than in less developed ones, but the same logic still applies: AI-driven software development is here to stay, and refusing to become part of it will hurt your business.

I'm also pretty sure that investors will start demanding that their portfolio companies start to use AI-driven software development. The ones that are stubborn and refuse will probably have less pleasant conversations with their investors.

This is a revolution -- and, for many, still an untapped opportunity. Don't be left behind. Join it instead of watching from the sidelines.

Posted by Per Werngren on February 23, 2024