Where in the past coding had to be done by a professional coder, nowadays generative AI such as ChatGPT has made its entrance. At first glance you might wonder how it works but by thinking it through it might actually be very logical. In coding classes, the first thing you get taught is: search on Google. This basically means that about everything you need to know about coding is available on Google. Since generative AI, such as ChatGPT, just use every information that is available on the internet, it is very logical that it is a good coder as well.
However, using AI to write your code might actually get you into trouble. Therefore in the work field, people use the AI to optimize their codes. This actually reduces a lot of time that can be used differently by professional coders.
This innovation can be seen as a disruptive innovation. Traditionally coders had to be extremely specialized in their work field, however by using generative AI, coding has become more available for a broader audience. It is also more cost efficient since labor where generative AI is used, relies less on expensive skilled labor. There are also some practical benefits of using AI for coding purposes. Generative AI is way faster than most of the humans in writing code and do it in the most efficient way possible.
However, next to the fact that it is a disruptive innovation, one should be careful with using generative AI for coding purposes. Does one understand how the code works that the AI has written? Or are we just living in good trust that no errors will occur in the future and if they do, we are able to resolve them? I guess we still need humans with specialized coding knowledge who can make notes in the code so we are actually able to understand how it works and anyone who is not familiar with the code can resolve issues that are popping up.
I like the topic you wrote about but I have to disagree with the argument made that ChatGPT uses everything you find on google. This is simple not true since ChatGPT 3.5 is trained with data sets from 2021 if i am not mistaken, hence it is fed certain information and others it is not. Likewise ChatGPT can produce useful and unusual codes, depending on its context, prime and prompt. I’d appreciate if you added a source to your document since this would have made it easier to follow your thoughts in writing this article. Furthermore, AI does lower the barriers of an average person being able to code but it is by far not reliable enough to replace coders. It will go a long way until this happens, which you also partially mention. However, it appears that you just scratched the surface of this topic rather than experiencing it yourself or delving deeper into it. An interesting take would have been the changes within the work environment for professional coders, how it changes their daily work life and how they can utilise this tool but also looking at its limitations. You also touch upon the issue of using AI to code but then continue with a completely different topic, I’d recommend you to structure your report better to provide a compelling argument. I’d like to engage in this article more in-depth but unfortunately, the information provided does not allow me too. I will share some sources which might be interesting to read!
https://olaf-thielke.medium.com/how-good-is-chatgpt-at-advanced-programming-5b8d75b45e9f
https://www.nature.com/articles/d41586-023-01833-0
https://analyticsindiamag.com/is-chatgpt-good-at-coding/
https://www.sciencedirect.com/science/article/pii/S2772485923000066
https://www.sciencedirect.com/science/article/pii/S2949882123000051