I have started working on a new project recently. This project has a lot of AI elements in it. Initially, it will not be a feature in the product, but gradually it will be there when the end product is reached. Since it is a new development, I wanted to use GPT for all my questions and interactions. It has been a great experience so far. I spend a lot of time with AI, and now I can understand what I should ask AI and what I shouldn't.
Sometimes the responses are amazing and very effective. If I ask the AI to give me an answer for a question, it gives me the answer and also predicts what my next question will be and provides me answer for the second question too. This is a wonderful thing. Sometimes it is very useful, and there are also times when I find it unnecessary and I ignore them if something like that happens.
This is something that I wanted to share. After using AI for so many days, even if I have the skillset to write some code, it is very tempting to send a prompt and get the code from the AI. The only thing that I have to do is review the code and make use of it in my project. This is very handy, and it is good because it gives me the best outcome code, and at the same time it is also time efficient.
Yesterday I had a problem to solve, and instead of going to Google search, I was using AI to get the answers for me and provide me with suggestions based on my code. I did share my entire code with the GPT, and it provided me with a lot of improvement opportunities. One thing to note here is that it can provide some sort of solution, but we have to validate it to make sure the solution is intact. Sometimes I end up catching some issues in the code, and when I get back to GPT about it, it realizes the issue and provides me with an updated solution.
From my experience, I would say that all the solution we get from GPT, we have to review and make sure it is not have any issues. If we don't care much about reviewing the code, we might end up in a big problem. I have had instances in the last week where I got a lot of wrong answers to the questions. For some answers, if I go and ask the GPT to review the answer it provided, it would then provide me a different solution.
From all these, what we have to understand is that AI can make mistakes and it can even make bigger mistakes, we should make sure the world is protected but doing a lot of review to the outcomes generated by AI.
If you like what I'm doing on Hive, you can vote me as a witness with the links below.
@balaz as a Hive Witness 
@kanibot as a Hive Engine Witness|
|
|
|
