GitHub copilot - Big Pros vs Big Cons
I started writing this article, three months ago but never got to finish it but now it seems more important to share this
I was one of the first people to sign up for GitHub co-pilot trial version, I was impressed with it. It is the first AI programming tool I used and it improved my efficiency by say 2% – 7%.
When the trial period was over and co-pilot launched the paid version, I was skeptical about if I needed co-pilot, I wasn’t sure if the increased efficiency was something I missed, or if it affected my work much so I opted not to pay at the time.
A couple of months after I got heavily involved in a couple of software projects and I began missing the hints and the extra suggestions co-pilot offered, now I needed all the assistance I could use to improve my efficiency. — My thought was if I could get say 2–3% increased efficiency in one project, spread across the four—six projects I worked on it would make a significant difference. So I subscribed for the paid version 😄.
Frankly, my second stint with co-pilot was even better, It greatly improved my efficiency this time I would say 7–12% this time. To be fair I think co-pilot was more helpful this time because I had been reading a lot of technical books on Software Design, Architecture, and Algorithms. — That was when I wanted to write this article, and somehow in between all the projects I have not been able to move past the title of this article, so here we are 😅 :
My thought and verdict on this, taking the current tech/software development ecosystem and the future of tech and software development into consideration.
GitHub Co-Pilot is an amazing tool, what is more, is that GitHub also announced GitHub Co-Pilot Enterprise license that in summary helps enterprises write code ( — ie generate code ) and as such
- GitHub Co-Pilot (AI tools) will greatly improve software developers' efficiency and this will make good developers better. It can make a significant highlight of a developer's strength
- On the organizational level, With better developers and even better tooling, organizations will need fewer developers to achieve more.
The negatives:
- There will be more gap between good developers and bad developers, great tools in the hands of good people make them better people. — We might argue that not-so-good developers also have access to these same tools, correct, but part of what makes a good developer is knowing the available options and the right option to use in an instance (use case) and why that is the best option
- There will be even more layoff developers because again, organizations will need fewer good developers to achieve more.
A side effect I think about this for developers is; software engineers will slowly evolve into software architects because AI tools will help generate codes and the engineers' role will be to review, and design solutions & strategies for functionalities
These are my thoughts on GitHub Copilot as of today. Of course, this is also true for other AI programming tools.
Look out devs 😉