How I've used gen AI / LLM
I write and design everything on this blog.1 Nothing is generated.
This is where I play, and I'm not letting ✨ AI ✨ do it for me. I like vexing over whether my use of a semicolon is correct (likely not) or whether a word means what I think it means (is vexing the right word for this?). I like writing verbose CSS and spending hours pruning it. I like the countless small choices I encounter when I create — the picking and picking and picking. It's fun for me. This is why I'm here.
—
I've used AI as a secondary tool for learning and problem-solving technical bits. When I couldn't find answers via search, I asked AI.
Examples of questions I've asked:
- When I didn't understand a command, script, or error code, I asked for an explanation
- While I navigated a problem, I asked if my approach would work
- When I wrote a command or script I was less familiar with, I asked if what I wrote had errors or how it could be improved
—
I once used AI to generate a programming function. I was not familiar enough with the language to know if it was well-written. I couldn’t be responsible for it since I didn’t understand it, so I scrapped it.
—
I've used AI to review my writing for clarity. I didn't care for its suggestions, so I stopped.
—
I used ChatGPT because it was the only AI tool I knew about. Then I saw multiple people mention Claude. If I use an AI tool again, I'd give Claude a try.
—
I'm subscribed to the News Minimalist newsletter, a news aggregator that uses ChatGPT to rank news by significance.
—
I experience cognitive dissonance over my usage, and I haven't reconciled it yet.
—
(I've just dumped this all here for now. I'll get around to organizing it later, maybe.)
{{ post_last_modified }} ago
Or copy & pasted it from others, in which case, they're referenced and linked. ↩