On 23rd March 2023 Microsoft had the coding world in the palm of its hand, while the world was still in a buzz about the newly announced GPT-4, Microsoft announced and gave a preview of Copilot X. Its predecessor Copilot is already a pre-existing AI-based code completion tool developed by GitHub and powered by ChatGPT. It uses GPT technology to suggest code snippets and complete code based on natural language inputs. Copilot was first announced in March of 2021 and now two years later it has now evolved to Copilot X which will be utilizing the more talked about GPT-4.
We are going to discuss the many new features it offers compared to its previous iteration and in what direction the programming world is moving towards with the integration of various AI-based code ‘Copilots’.
Literally a Copilot: Chat Window and IDE integration
Copilot X is specifically designed and optimized for IDE integration and can access your workspace and get the whole context of it and its Chat Window acts as an assistant that can provide insightful suggestions, highlight errors and even make the code more optimized. You can highlight a block of code and can enter prompts to change the code and replace it with the AI-generated one. Although it can be done with ChatGPT what makes a difference here is the GPT-4’s 25000 token limit, compared to the 3000 token limit of GPT3.5. With a capacity of more than 8 times and optimization for several IDEs including JetBrains. Say hello to your new coding partner.
Documentation and implementation made easy:
Remember what I said about the advantage of GPT-4’s 25000 token limits, turns out GitHub is integrating the Microsoft, React, and Azure docs in the Copilot X model. Whereby you will be able to get answers to all your queries when you are new on a platform and don’t understand the functionality just yet. Gone are the days when you had to go through paragraphs of documentation before you start working on a new project and since the docs are integrated in the model a lot of well-versed solutions can be provided to the users.
Voice Coding: “Give me a hand with this Jarvis”
Before talking about this one, I believe this is the next stepping stone of what programming is going to look like in the next 10 or maybe 8 years where a lot of work is focused on logic building and project designing and full automation of tedious code writing. We already use a mild version of it without knowing. We have already been using the code completion feature of VS code and other IDEs and it has been so integrated into our lifestyle we cannot imagine us going back to the days of typing each code line without these now fundamental features.
Voice Coding aims to take this to another level of sophistication. It does exactly what you think you can now code with voice commands, a feature that is so simple but took years to implement is going to revolutionize how most technical spaces work and how even a general programmer sitting in his room codes. Imagine being able to code on the fly, wherever you are whenever you are. Being stuck in traffic doesn’t matter just voice what logic you want to implement. This is the one feature that I can’t wait to become a part of our routine, the same way Google Home and Alexa have become a part of our lifestyle.
Less time spent on Pull Requests Description:
The two hardest things in programming are naming a variable and documentation. Copilot X comes again as a lifesaver. It is now capable of automatically analyzing your code and the changes you made and in accordance with that can produce concise PR descriptions. Copilot just makes the process smoother and more efficient indirectly reducing the time a person spends on writing PR texts and not having a breakdown thinking no one is going to understand anyway what they did.
Support for CLI: It keeps getting better.
The auto-completion feature is not only limited to the text editor but also the terminal commands. The terminal is a kind of place that has a huge arsenal of capabilities with which it completely changes how we write code or even interact with our computers. But it does come at a cost and takes many years of regular just to become proficient at it (Looking at you Linux).
Copilot X, more like Shell Wizard X, which already has accumulated all the knowledge that you need to know of. All that is just at the tip of your fingers. Just type in the library name and the command you want to generate and your wizard model will generate the terminal code along with the explanation, again emphasizing the fact the future of programming is centered more around logical thinking and less about looking at the abstract details which the AI is going to implement for you anyways.
A final Compilation:
With such a massive and perplexing technology just around the corner, it also begs a massive question. How is it going to change the scenario of coders and the general job market? This a question whose answer looks threatening at first, but if we dive deep between the lines it going to completely change how an average programmer generally codes. A lot of the future is going to be emphasized on thinking about the fundamental concepts and just making the programmer more efficient. It does so by reducing the time we spend on redundant tasks which will now be handled by AI models. I don’t know if I am just looking at the silver lining here, but this all just makes me very excited about the future we are headed to. Even 20 years ago people knew AI was the future and now while we are at it is the right time to incorporate that to our work routine, even though I have said it all along before. Copilot X is for drastically raising the efficiency of the entire coding world.
So, if you are also eager to try and get an experience of the GitHub Copilot X, sign up for a technical preview on the GitHub Copilot X website and aboard the hype train. Hope you have a great time coding.
Top comments (0)