How to Use Chat Gpt Right Way as Developer
Chat GPT is like machine gun, it can fire a lot of bullets, but its equally hard to aim right. If you really know what you are doing, its better to use a sniper rifle. But if you are not sure, its better to use a machine gun.
If you ask Chat GPT to write code, it takes 1 minute to write code, 1 hour to debug it and fix it if it did something on odd line you didn’t notice or language you don’t understand. Writing code line by line helps you understand, you take time to read documentation of each new function or class and change it better. so, writing code which is not trivial is not a good idea.
But it is better to search ideas but it requires some skills in prompting like give me ideas on how to design blog database is not much useful but if you build context like example below.
You are software engineer, you are tasked to design a blogging which should be SEO friendly, should be sharable on social media. It should be cheap to deploy and can scale up to 100000 users per day. It should be easy to add content and should be easy to search. Write a design about what databases will you use, compare multiple options and pick one.
It can search technologies or ideas you have never thought and knew before. It can replace google but not your development skills.
Another way chat gpt really shines is data transformation. Suppose you have C#/Java class and you want to convert it into SQL table or CSV data to JSON.
It also helps building queries, like you have a table and you want to get data from it, you can ask it to write a query for you. Even AWS cloudwatch insights query can be written by it.
Other than code, It also helps in repharasing simple english paragraphs, professionalizing emails etc.
But don’t get too dependent on it writing code, it can hallucinate functions which don’t exist, write terrible code which you will have to debug for hours.