Frustration.
It’s one word that comes to mind when you ask ChatGPT about something it doesn’t know, only for it to run in circles for hours giving you non-answers and inviting you to use libraries that do not exist.

When ChatGPT first came out, many of us initially viewed it as an incredibly impressive tool, one that had the potential to replace even an engineer someday. However, after using it for a while, I’ve realized that the above statement is largely inaccurate. While I’ve had positive experiences with ChatGPT, it has also provided me with questionable coding advice and generated graphics that wouldn’t look out of place on a website from the early 2000s.

Now, let me clarify something. I do recognize the significance of ChatGPT. However, it’s not significant enough to fully replace Google or other comprehensive search engines. There have been instances where I had to resort to extensive Google searches to find information on enabling ffmpeg on Android with OpenCV because ChatGPT’s answers were consistently wrong and trapped in an unhelpful loop.

That being said, ChatGPT has indeed assisted us in setting up Gammu, a valuable tool for mobile communication. It has also been useful in providing recommendations for libraries that can help you get started on a particular task. While ChatGPT may lack the deeper understanding required to implement these libraries, it can serve as a starting point and offer valuable suggestions.

Additionally, one aspect of ChatGPT that can initially appear impressive is its ability to provide information on seemingly obscure or niche topics. It can give the impression that it has extensive knowledge in various domains. However, upon closer examination, it becomes evident that ChatGPT’s responses often tend to be shallow and lack the depth of understanding that is required for meaningful insights.

Furthermore, one of the challenges of relying on ChatGPT for coding-related inquiries is the potential for misleading or incomplete information. While it can provide recommendations for libraries and offer suggestions, it might not possess the practical understanding necessary to implement those suggestions effectively.

Coding is a complex and intricate discipline that requires a deep understanding of programming languages, frameworks, and best practices. While ChatGPT can assist in pointing you in the right direction, it’s important to approach its recommendations with caution and cross-reference them with reliable sources.

Moreover, ChatGPT’s limitations in generating graphics can be another source of frustration. Although it can generate visual elements, the quality and aesthetic appeal may not meet modern standards. Its output might resemble graphics from a bygone era, which can be discouraging for those seeking sleek and visually appealing designs.

To navigate these pitfalls, it is essential to approach ChatGPT as a tool to augment your knowledge and aid in initial exploration. Supplement its responses with thorough research, documentation, and input from experienced developers. Engaging in active learning, seeking guidance from coding communities, and consulting reputable resources can help fill in the gaps left by ChatGPT’s limitations.