unknown

AI-driven tools like GitHub Copilot are revolutionizing programming by generating code and suggesting solutions, but they may lead to an erosion of core programming skills, over-reliance on auto-generated code, reduced responsibility, and fewer learning opportunities. These tools could also limit creative thinking and create dependency on proprietary platforms, potentially giving programmers a false sense of their expertise.

Why Copilot is Making Programmers Worse at Programming

Squad for devs building apps with GenAI and LLMs! Connect, share, exchange ideas, and get feedback. 

Build With GenAI

Why Copilot is Making Programmers Worse at Programming

I strongly disagree with the article.
You are not supposed to use AI for everything. Moreover, AI is not good at doing these things.
<ul>
<li>AI is great at generating boilerplate code or some things that you know how to do, but you don’t want to bother to remember the syntax.</li>
<li>AI is good at generating UnitTests.</li>
<li>AI is good at documenting your code.</li>
<li>AI is good at helping you understand better the codebase of your project.</li>
<li>AI is good at brainstorming ideas on how to approach a difficult problem</li>
</ul>
On other hand
<ul>
<li>AI is not good at generating ready-to-use complex algorithms.</li>
<li>AI is not good at generating large applications or features that don’t have specific requirements</li>
<li>AI is not good at integrating code inside your codebase, following the conventions and best practices for the project.</li>
</ul>
I’d say that people who are trying to delegate all of their work to AI, either are not producing good results, or they are just not needed.
AI is here to make engineers more productive. None would say that IDEs are making programmers worse at programming, or autocomplete, or personal computers… Let’s get back to writing code at punchcards and be great programmers again.