Why Github Copilot is bad for a programmer's health

Isaac
3 min readJul 13, 2022

I was a day one subscriber for the Github copilot beta test and truthfully I loved it, feared it, and ultimately terrified of it. Heres why…

Blazing fast coding

Like most programmers, you want to use the latest tools, libraries, and APIs to help improve your coding project and performance. Some applications do just that and can reduce a headache when you encounter a complex problem. Then came along your copilot. “Okay, an AI assistance. What can go wrong?” Well depending on what language you program in, GitHub Copilot can provide inline coding suggestions, as well as comment request suggestions.

For example:

// function that finds the k-nearest neighbor
(alt-enter)
Comment on Left side, Suggestion (alt-enter) on right side.

As shown above, the code provides a suggested code snippet on a new tab. Fine so we accept the first solution out of 10 possible suggested solutions.

Done! All we need to do is maybe tweak the parameters or even rewrite the first line then ask Copilot for suggestions after that, there are all sorts of approaches you can do as long if it has sufficient data prior to the AI search.

Beware of the “Fog”

I can honestly say that my productiveness increased by 100% or doubled by cutting my coding time in half or more. It was great and life couldn't get much better, and the future started to look a lot brighter.

One year later, I’m still using GitHub Copilot and it appears my dependency on the AI assistant has sunken its claws. I use it on a daily basis including for my personal projects. Big deal, right?

I’ve recently come into a dilemma where I’m solving problems on HackerRank, a third-party problem-solving application, (No resources allowed) just you and a keyboard. “Okay I got this”

Wrong! I’ve come to the realization that most of my coding experience still exists somewhere in my brain but have developed a co-dependence (no pun intended) that accessing that information seems to be redundant.

Given, that you may not experience the same situation as I have but will realize a part of you is missing, and I don't mean it in that sweet emotional way. I mean in the, “Where am I?” way. It appears that the lack of need to critically think of lower-level problems forces the overly ambitious mind to look for new abstract problems perhaps higher-level at that. Which is great if you see the long-term aspect of it, though was the trade-off worth it? I believe only time will tell, and that I will probably never go back to the old way of programming. That being said, will I be able to program without Copilot? Sure, but with the same enthusiasm as remembering your friend's new phone number by memory.

I’d like to explain what it feels like more in-depth, though I could probably leave you with a funny meme to ponder.

To point out, I am currently a Software Developer and broadening my scope to Machine Learning and AI. Please clap if you enjoyed my topic and leave a comment.

--

--

Isaac

M.S. Electrical Engineer, Working as a Software Developer mainly with AWS.