Hello and welcome to the channel. I recently made a video about the advantages that Python and other popular programming languages have over click and drop software packages for data analysis. The link to that video is given in the description box.
In one video, I have shown how you can use JTRPT to write your codes, which you can then simply copy and paste into an IDE, for example, R Studio, and do your analysis. You can do the same in Python as well, by asking Jet GPT to write a code for the task you want to complete, and then simply copying that code and pasting it into an IDE of your choice, for example, Jupiter Notebook.
In another video, I have shown how you can use Jet GPT’s recently released code interpreter plugin to do analysis with ChatGPT, even without having to copy the code generated by ChatGPT and pasting it into an IDE. With that plugin, ChatGPT can do all the analysis for you and give you the output.
The question is, if ChatGPT has this capability, is it still necessary to learn R and Python? The short answer is yes, you still do. There are many reasons for that.
Firstly, ChatGPT is not perfect. It still makes mistakes in the code, and sometimes when you copy the code and paste it into an IDE and try to run it, you get an error. Similarly, it is quite possible that it makes a mistake while you are using the code interpreter plugin. So, you cannot totally rely on ChatGPT. You must know the language and have some understanding of these languages to identify and correct those mistakes.
Secondly, ChatGPT has a knowledge cut-off date. It has no knowledge beyond September 2021. If new packages or techniques have been released after that date, ChatGPT does not know about them. So, its knowledge is not up to date with the latest developments.
Thirdly, ChatGPT has limited capacity. Even with the code interpreter plugin, if your data is too large, it may encounter an error and be unable to process the data. In such cases, it is more suitable for you to do the analysis yourself.
Fourthly, ChatGPT and its code interpreter are unable to access the internet directly. For some tasks, if you give that task to the code interpreter, it will need to download a package or library. But it does not have the capability to go to the internet and download those packages required for certain analysis. In that case, it will not be able to help you, but it can tell you which packages you can download.
Fifthly, ChatGPT is an AI language model. It is not a human brain and lacks the creativity of a human brain. Sometimes, you will encounter a unique problem that requires a creative solution, and ChatGPT probably cannot provide that. Only you can do that. If you do not know how to code, you will not be able to do that kind of creative analysis.
Lastly, ChatGPT lacks domain knowledge. It can give you the code, but it cannot interpret the results for you because it may not know about the specific domain you are conducting the analysis in. The interpretation of the output of a data analysis process requires domain knowledge.
In conclusion, the right use of AI tools like ChatGPT is to increase your productivity. It is not recommended to become dependent on AI. Instead, you should use AI to increase your productivity, for example, by asking it to generate an initial code for you and then fine-tuning it. However, if you become too dependent on AI, you may become less creative, and AI may replace you. Therefore, it is important to learn programming languages and stay ahead of AI to survive in the AI age.
Thank you.
