Artificial intelligence and machine learning continue to prevail in the modern age. Most of us trust these technologies in some way, and GitHub developers certainly fall into that category. Since the website’s Copilot tool launched in June, almost 30 percent of the new code has been written using AI.
If you’ve never heard of Copilot, it’s essentially the equivalent of the predictive text systems you can find in Gmail or even on your phone. Start by writing a line of code and Copilot can auto complete it for you or try anyway.
Surprisingly, more than 50 percent of programmers who chose to use Copilot kept using Copilot after the first use. We don’t know how long they’ll keep using it, but it’s still a good sign when such a large segment of GitHub users don’t just scrap the technology after the first try.
“We hear a lot from our users that their coding practices with Copilot have changed,” said Oege de Moor, vice president of the Copilot team at GitHub Next. “In general, they can be much more productive at coding.”
It’s not difficult to see why that is. The pilot can autocomplete large chunks of code, and while there are likely to be bugs here and there, it still seems like a tremendous time saver in the long run. a programmer to write standard repetitive code and allow him to focus more on polishing.
Currently, the GitHub Copilot feature is only available for a handful of programming languages, including Ruby, TypeScript, and JavaScript. Moving forward, de Moor and co. plans to add support for additional languages, or at least the most popular ones on the market. The feature isn’t perfect, of course, but the more users use it, the more sophisticated it gets over time.
If you are a programmer who has used Copilot in the past, feel free to leave your thoughts in the comments below. We’d love to hear some firsthand reports about the accuracy and efficiency of the technology.
