By
Originally posted by Toni Piehl as LinkedIn article on October 18th 2024.
Generative AI is here
Many of us are wondering what think about generative AI, will it take our jobs and what to use generative AI for. So far the general opinion is that those using generative AI will remain and others are relatively loosing their career development to those being more productive with the aid of generative AI.
Use case for GenAI
I started looking at my LinkedIn background image yesterday after earlier having a question from a head hunter in meeting “Did you code the word cloud in your LinkedIn profile yourself?” and I had to say let him down and say I didn’t. I’m not a developer per se but, I thought it can’t that difficult nowadays. So I turned to ChatGPT 4o and gave it requirements for python code to draw word cloud. After more than a dozen iterations the code was working as expected. It took still some modifications to find a design I liked, and once the word cloud was ready I still was not happy with the way it looked. So I turned to ChatGPT again to draw comic style cloud I could use as a mask for the word cloud. I edited the provided image in gimp to match my need. In the end I am satisfied with the silver lining of the word cloud which is now my profile background. Let me know what you think!
Learnings from the exercise
– ChatGPT writes high quality code and it imported useful libraries on its own (versus having to code all of it from scratch)
– Written code was well commented, easy to read code
– Instructions for getting the code running were provided
– Not each code version worked. Main issues were handling of alpha channel with word cloud library (background transparency) but after providing Traceback, errors were resolved.
– I manually edited some code versions, but it was more time consuming than just asking for a new version with new requirements
Experience
ChatGPT was acting as a colleague which was tirelessly ready to hear new or changed requirements, listing wanted changes, explaining proposed solution, providing the python code, listing the adjustments made in relation to prior code version, explaining the expected output and asking for feedback whether code performed as expected or if any modifications were necessary.
Summary
The exercise gave me a good understanding that AI aided code development is the way to go even for occasional developers like myself. Especially trying to find out fixes for code errors is usually troublesome as it could be in the way imported libraries are used or in the developed code itself. Now errors were resolved by AI developer in fast and easy way. I will continue using AI in my occasional coding exercises.
#chatgpt is from OpenAI #generative #ai #generativeai #code #development
