I focus on the OpenAI chat models, which are optimised for people who just want to get on with it. To access them, you need to sign up, but you can use them for free or via a paid subscription. The benefits of the paid subscription include anytime access to the models (they are big and expensive to run, so if the servers are overwhelmed with users, free tier users are paused) and access to a wider range of models.

The first thing I did was to use the LLM (Large Language Model) to make my CV more professional. The message box, where you provide "prompts" to the model, doesn't allow for much formatting. This actually helps keep you on point because the models themselves do not understand formatting. If you rely on formatting for context, the model will struggle more than if you could provide all the context in plain text.  Despite this, I find it hard to cope without the ability to press "enter" to go to a new line. If you press "enter" while in the message box, your message will be sent to the model. For a newline, press "shift-enter."

Getting ChatGPT to make me more professional turned out to be much easier than actually getting qualifications or actually becoming more professional. All I did was enter, "Below is my CV. Please make it more professional." I then pressed "shift-enter," copied my CV from Word, and pasted it into the message box. Afterward, I pressed "enter" to send this message to GPT.

Note: There is a limit to how much information you can include in a message. More on this later, but for now, please note that if your CV is full of accomplishments, you may not be able to paste it in one go. Fortunately, my list of accomplishments is short.

ChatGPT created a much better version of my CV. I have discarded my original version and have kept the GPT one.

However, the ChatGPT version wasn't without its issues. In particular, it kept all my academic qualifications but changed the qualifying institution. I studied at the University of the Witwatersrand in Johannesburg. GPT gave me my degrees from Oxford. Now, while I fully accept that the degree from Oxford would have been more professional, I'm not sure that the instruction to make me more professional should have been taken quite this far!

To conclude the CV process, I copied the text from GPT back into Word, added back formatting, and manually fixed any "upgrade" that GPT gave me that I felt might have been pushing the boundaries a bit. So, back to Wits it was for me.

In the following posts, I will explore GPT in more detail. But even after all the time I've spent with GPT, I still find the best use for it in my daily life is pretty much this example. I write something, I ask GPT to fix it (spelling, grammar, tone, and length), I then correct any issues it created, add back anything I really thought I needed but that GPT may have summarised out, and proceed with the result. The output is significantly better than what I could have achieved on my own. You should also try it – unless, of course, you're a language major or professional writer. GPT is good, but it will never improve Terry Pratchett or Tom Robbins.