07/06/2021
Just waiting on GitHub Copilot to magically finish my feature for me 😋
For those who missed Microsoft and OpenAI’s recent collaboration, a technical preview of GitHub Copilot, “a new AI pair programmer that helps you write better code” was launched last week. —Nat Friedman
As a majority of us are still working from home, this could be just what the dev community needs. GitHub Copilot could prove to be an effective solution for more streamlined, efficient code!
Game changer. With endless possibilities from suggesting entire functions to generating Functional API keys, I can’t help but wonder:
👉 Will my “Copilot” improve my coding skills when learning a new technology/language? Beginner 🔜 Advanced
👉 With the ability to code faster, will the pressure be on for developers to push out features at unsustainable rates? And if so, how will it affect the integrity of our code? 📈
👉 Will it become standard practice to use GitHub Copilot as a preemptive, automated Code Review?
👉 Will SDEs thrive with an even more lax approach to coding, or will we be rendered less valuable overtime as programs and products are built based on context and AI, thus lulling the high demand for more developers?
👉 Will GitHub Copilot provide answers that Google and Stack Overflow can’t?
I am positive testing will be a must. While GitHub Copilot has the ability to recommend and even write tests, our cybernetic sidekick does not test the code it suggests beforehand. This can result in errors, edge cases, failure to compile and code that simply doesn’t work.
I’ll keep you posted on pros, cons, use cases, etc once I’m able to explore it a bit, but with the waitlist for the technical preview it could be awhile 🤷♀️
✨What are your thoughts on AI Pair Programming? Could this be a fundamental shift in the way we code and the future of the high-tech industry?