AI use cases for GPT and more

The image to the left here is generated by midjourney, but this article is about (chat)GPT and how I've become so much more productive thanks to it already.

1. Thinking through a project idea. I wasn't excited about my work project, but GPT took away the boring stuff. Consider:
   1. Competition research / SWOT analysis
   2. Consider the risks and challenges of scaling this specific project
   3. Defining a mission statement on the learnings thus far
   4. Making job descriptions for hiring the right people

ChatGPT does some amazing 'boring work', perhaps even 80% of it. I'm fascinated!

2. Designing a garden based on research done by GPT4
   1. Using that input for a mid-journey prompt
   2. Finetuning and being amazed by the amount of creative input these models generate
3. Thinking through a data structure for a table, I wanted to create and give relevant and usable input to improve. I can only imagine what it can do for database normalisation or API designs.
4. Lots and lots of code samples for javascript, python, CSS, HTML, you name it. Even putting it together in one!
   1. Most recently, using GPT as a 'code assistant' to validate and simplify some complex functions in node-red.
5. Figuring out how to extract text from a PDF using a Python script (without former Python knowledge)
6. Learning new terminal commands for my Ubuntu VMs
   1. Finding the largest file in a folder

      find / -type f -exec du -Sh {} + | sort -rh | head -n 1

   2. Echoing text into a file (GPT also gave me the option to append or overwrite!)

      echo "This text will be added to the end of the file." >> filename.txt

   3. a CURL request with a POST body

   curl -X POST -H "Content-Type: application/json" -d '{"text":"Hello, world!"}' http://basic:auth@domain/gpt3/

7. Taking a press release and making it my own
   1. translated it to English (from Dutch)
   2. made an executive summary
   3. turned it into a haiku
   4. rewrote it from the perspective of a consultant
   5. Generating social media posts out of it (in different languages)
8. Doing social media posts for my blog based on the content. This probably will be the death of social media (again).
9. Comparing a job offer to the demands I have of a job. Being on top of market demands and managing your network and connections are all part of the deal.
10. Generating a test in CypressJS (my first task)
11. Using chatGPT to read, summarise and enhance my texts based on the knowledge it contains. Being a trustworthy 'assistant'.
12. Preparing myself for meetings by letting chatGPT challenge my assumptions and agenda

1. Google.
2. Being annoyed by stack overflow
3. Spending energy on seeing (or blocking) ads
4. not hoping the code works that I copied from stack overflow, but knowing why it works.
5. data entry, analysis and research

That it's not rocket science, it's not insane. It is understandable, manageable, and straightforward. And that's what is accelerating this. There is acceptance for the 'bot' part (thank Siri/Alexa for that). There finally is a better Google.

No, scratch that. Ignore that. It might be Google. It might be a junior bookkeeper or senior developer. Accept that it is an assistant. An entity that helps you to learn to become a better (coder/writer/researcher/creative/fill-in-the-blank). It is biased; be aware of that, but it will improve your work.

The decision is not (yet) out, but there are some ideas about where the rights lie.

In short, you own the input, and they own the output.

• https://www.jdsupra.com/legalnews/chatgpt-who-owns-the-content-generated-2891692/

There is a fear that the data we enter into any learning model will be used to improve the model itself. I don't have anything against that in particular, but if you don't want that to happen – make sure you fill out the following form(s):

• OpenAI Data Opt Out Request (ChatGPT, DALL-E)