A little back story… for my LLC Mavisworks, I have thrown myself back into hands-on technical, which I deeply miss and enjoy. For me, working on websites has always been nerd fun. The 2 apps that I’ve used more than others are SharePoint and WordPress. I like having the ability to use the MS stack, but enjoy working on my Mac and building out sites. I have a long-term relationship with my host, A2, and own several domains.
MVP for mavisworks.com was to purchase the domain, throw up a placeholder site, secure it with SSL, and have my business own it. It took a bit longer than I would have liked given the challenges of setting up a new business while simultaneously building infrastructure.
V1 is live and is reasonably ok. I have a product backlog of work that will keep me going for a while. V2 will fix some additional css and add a password protected portfolio subsite.
Anyhow, in troubleshooting today, I started doing web searches and was not getting very effective help. Given my history with building and implementing search, that says a lot. I opened up ChatGPT and had a conversation about the issues that I was working on (main logo and alt logo sized differently and very distracting) and got some really good ideas and information on how to resolve my challenge. For gen AI, this iterative prompt refinement is called Chain-of-thought prompting, which helps both me and my AI chum to work through a several step, or complex question. Or additionally, I likely needed to refine the language I was using in my request.
One of the things that I am concerned with gen AI is generating content from legitimate and reputable resources that are cited. I got exactly that in a display of resources and answers.
I’ve saved the conversation for future reference. My ability to ask questions, including “why would anyone want to increase the size of a logo on scroll,” got me a combination of current best practices with specific css and js files for my particular install of WordPress and accompanying themes and plugins.
It was both entertaining and efficient, as it took me less time than I anticipated to figure out what I wanted to do and “get ‘er done.”
Have you had satisfying experiences with gen AI? Do you have a favorite LLM?
Comments are closed