GitHub’s Copilot: The AI Revolutionizing the Coding World
Over the last decade and a half, GitHub has become an essential part of the coding sphere. Launched in 2007, this platform is now relied upon by over 100 million developers to collaborate, track changes, and store their code.
Fast forward to 2018, and Microsoft acquires GitHub for a whopping $7.5 billion. Is that the end of the story? Oh no, not even close. GitHub is now making waves for another reason: it’s transforming the way coding works, all thanks to Copilot, an AI tool that generates code as smoothly as OpenAI’s ChatGPT produces poetry. (Fun fact: Copilot is powered by OpenAI’s Codex, which is a large language model trained on code collected from the internet.)
Making Coding a Breeze with Copilot
By simply hitting the tab key, a developer using Copilot can finish a line, generate blocks of code, or even write entire programs. According to GitHub, more than 10,000 organizations (from Coca-Cola to Airbnb) have signed up for Copilot’s enterprise version, and over 30,000 employees at Microsoft now code with assistance from Copilot.
Freethink recently spoke with Thomas Dohmke, GitHub’s CEO, to learn more about how Copilot is reshaping the programming profession and the questions AI-powered development raises about the future of innovation.
The Developer’s New Skills: Leveraging AI
Dohmke believes that systems thinking and breaking down complex problems into smaller ones will play an increasingly important role in software development. He emphasizes that developers will need to figure out how small a problem should be in order to leverage AI for synthesizing code.
However, developers will still be the experts and will need to understand and review the code synthesized by AI to ensure it matches their intent. The balance between human-written code and synthesized code will likely shift over the next decade, but Dohmke doesn’t think no-code development is taking over anytime soon.
Day-to-Day Changes for Programmers with Copilot
With Copilot, developers can make better use of their coding time (usually only two to four hours a day) and stay in the flow. The tool helps reduce distractions and allows developers to focus on creativity and problem-solving.
When asked about his own coding process with Copilot, Dohmke shared that it brings back the fun and creativity in programming. The tool helps him stay in the flow and enjoy working on his hobby projects.
Copilot: A Game Changer for Startups and Established Developers
Dohmke believes that Copilot would have made his life as a startup founder much easier, as it helps maintain the flow and quickly reach prototype and product-market-fit stages. Many startups and angel investors he knows are using Copilot for its productivity gains, which can be crucial for their survival.
When Copilot was first tested internally, the response was overwhelmingly positive. The net promoter score (NPS) from testers was +73, an impressive number for a new tool. Retention numbers were also higher than any other product they had built.
Copilot’s Impact on Productivity and Learning
On average, 46% of the code in files where Copilot is enabled is written by the AI. As a result, developers can focus more on the remaining code they need to write. Copilot also changes the way developers write code, as they become more explicit with their intentions.
A study found that, with Copilot, developers were 55% faster at building a web server and had a higher success rate (78% compared to 70% for those without Copilot).
Dohmke believes that Copilot will change the way programming is taught and learned. With AI tools like Copilot, students can interact with programming concepts in their native language, making the learning process more democratized and accessible.
Open Source and Copilot: A Powerful Combination
Open source plays a fundamental role in the evolution of Copilot. The AI model is trained on open data from the internet, including source code and other documents. This helps Copilot understand not only programming languages but also human languages.
GitHub has always supported the open-source community and will continue to do so. Dohmke believes that large language models will significantly contribute to accelerating progress in the field.
Copilot and the Future of Programming Languages
Languages may go in and out of fashion, but they never truly disappear (COBOL, anyone?). Copilot can help modernize code bases and bridge the gap between old and new programming languages.
As AI tools like Copilot improve, coding may become more similar to human-to-human communication. Developers will be able to interact with AI in a more natural way, making the coding process more intuitive and enjoyable.
Overall, Copilot is a revolutionary tool that is transforming the coding landscape. By enhancing productivity, reshaping the learning process, and redefining the collaboration between humans and AI, Copilot is paving the way for a new era of innovation in the world of programming.
