Experimenting with AI

I have made no secret of my displeasure with a lot of the hype and uses for AI. I still feel that way about a number of things that AI is used for, but I've also found some areas where it has shown some legitimate use-cases (the automated code review we're using at work as part of a larger, human-involved code review process frankly blew me away). As a result, I'm trying to stay open-minded and test out various scenarios where it might be beneficial to me.

I created a small Gemini gem to automate the process of making book quotes. My rationale here is that the alternative is my manually creating them in an image editor (and I never feel like going through all the effort). It's not taking away the work of a human except my own, and the artistic expression is about as much as I want.

I really wanted to automatically include the cover of the book, but the current version of Gemini doesn't have that ability - I'd have to upload the cover manually, which is the sort of effort I'm trying to avoid. But it did offer to try to recreate covers manually for me, and then use the accompanying representation.

I'm not gonna lie, I almost went for it. It's close enough to impart the general idea of the cover, and for a minute I was even tickled by the idea of having custom covers for the books, similar to the generated UX I keep hearing lurks just minutes away from sweeping away all design, ever.

But upon reflect, that's just a bit too far for me. Those covers were (hopefully) made by humans with an artistic eye and vision they were trying to impart, and to replace it with an AI imitation is to cheapen their work and lessen the impact of it.

It is, however, super annoying that this version has a much better layout overall.