Working with AI

I’ve been working on some AI shenanigans lately, largely finding way to leverage GPT’s power to benefit myself both privately and professionally. Since I started using the venerable AI, I have attempted several, to some small degree of success, projects. Most of them are writing prompts that have led me to some really wonderful creative exercises and experiences so I’m really grateful for that. But in the realm of technology, GPT has helped me brush up on some html and python… and that’s about it.

I can see how being a visionary or creative can be wonderfully addictive and easy, at least in the sense of putting the rough draft together and sending it off into to a workshop to get put together, then launched in whatever capacity it shows up as. I mean I got that down pat. Need an idea? I’m your guy. Figure out how to execute? I’m your guy! Put it together on a PowerPoint? I’m your guy!

Actually execute the backend code?

Print<WOAH>

My experience with Python was rudimentary as part of a CS class. With HTML, its kind of archaic. But I want to keep expanding my horizons and work on these two languages and understand how to really manipulate them to my advantage. Much like the English language, which at times I command, and others, fail at, it will allow me to communicate with electronic tools or show something visually (or technically) stimulating to my audience.

So, my current conundrum with GPT is that while I can give it very specific directions, the program will sometimes decide to do its own thing. Think of it like this. I tell my son to go straight as we head to the park. The park is the goal. The steps and the distance is what we have to overcome to get there. Its literally “Son, follow the path, one step at a time, don’t trip over yourself, and get to the park”. I turn my head for a second and the boy has somehow made it into the street to oncoming traffic (Don’t worry, this is an analogy – my son is usually smarter than that… usually). Regardless, I find myself having to go back and adjust both the AI’s results and temper my own expectations.

Don’t get me wrong, GPT and the derivatives around it are some really fascinating programming and have a ton of potential for some incredible usage. Yet AI is still in its infancy and while can course correct after the fact, in the moment, it behaves almost like our children do.

Imagine that.

I can’t wait to dream in code.

Leave a comment

Your email address will not be published. Required fields are marked *