I'm glad you enjoyed the game.
I appreciate your thoughtful response about the music and sharing your feelings. I knew being upfront about it could potentially tank my ratings. The indie game community is pretty anti generative-AI and I know this. In fact I used to be rabidly so. Then I ran into a cute little game with really good music in a Jam. In fact, it was the GameDev.tv Tiny World jam. So I decided to try it. I speak about this in my Eternal Echoes Pots Mortem DevLog. In there, I also talk about my mixed feelings. I've also had some very long discussions about it in the Godot forums. Including this week, an off-topic reply about generative-AI that sparked interesting discussion.
I don't want to put artists out of work. In this game jam, no musician lost out on money because I couldn't spend money on them to make music. Could I have found a music producer to do this for free in a week and make the song I envisioned, as well as two chiptune versions of the same track and a fourth version of the track with Skele-Tom humming it to himself? Maybe. My brother is a DJ and music producer, and we didn't have the time to do it together. And the fact of the matter is while the generative AI didn't perfectly recreate my vision, it was close enough for a game jam.
Sure, the generative-AI did a lot of work in literally seconds. But it wasn't like I took the first version it spit out. I wrote the lyrics to have a particular feel. I asked for a particular type of song, and it took a number of tries of me tweaking the lyrics and prompts to get what I wanted. I believe it was the 8th version I picked. I do not think I could have done that without my own (limited) musical training. I definitely know it wouldn't have been a banger if I'd chosen the first version. I literally tweaked the number of syllables in the verses and chorus to make it flow perfectly.
I've read discussions on music forums about generative-AI in general and Suno (the tool I used) in particular. One example was a songwriter who uploaded him singing and playing guitar to a song he was working on, and then using Suno to add a backing band to him so he could see what it sounded like. does that take away from other musicians? Maybe. But I wonder if there's a world between AI-slop and "pure" music where these tools can be used as creative tools instead of creativity replacements. One of the (amusing to me) things I've seen on Suno is people copyrighting their lyrics in song descriptions. Which misses the point of how generative-AI steals from musicians to work.
It's an ethical question that I have been pondering. I thought the courts were going to shut it down with the MidJourney AI lawsuit. When they didn't, I realized it was here to stay. So I'm wondering is there way to use it ethically. There are for example, ethical implications to using LLM AIs to help out in school. But students don't seem to care based on what I read. So do I become an old person who just won't change on principle? I honestly still haven't decided.
However one feels about it - I couldn't have made this game without the music. It would not have been nearly as good. I listened to it all week, over and over as I was testing levels, and I never got tired of it. I could have probably found a banger chiptune track for free, but I couldn't have put the story of Skele-Tom into it and gotten the feel I wanted for the game. And yes, it is in one way upsetting that it's so good. In another, as someone who always wanted to write songs but had the poetry but not the musical knowledge (not for lack of attempting to learn), it's been really fun to write songs just for me as a way to process and express my own feelings. And I see a lot of people on Suno doing the same thing. Processing heartbreak, loss, hurt, frustration, and celebrating love, happiness, excitement.
I do have opinions, but they're not fully formed. And I appreciate thoughtful dialogue with people on both sides of issues like this.