Utilizing ElevenLabs opens up some possibilities for voice work that I otherwise haven't been able to achieve on my own and cannot afford to hire voice actresses for. In combination with using Koikatsu and Honey Select, and sound effects and samples from places like Freesound, I could theoretically make some mini-movies with full voice acting and minor animations and cuts in the style of the last video. For the videos above, the first three are using text-to-speech, while the fourth uses speech-to-speech.
That said, it is still a very frustrating and clunky technology to use, and the whole use of AI for creative work is something I do have a lot of mixed feelings on. I had initially posted a big fat opine about it, but considering the resources I've used in the past to make my "transformative work" in the form of sniping images from across the web to write captions for, and using Illusion Studio's games to make machinima comics, well, I'm probably not a very good person to try and be a paragon over the morality of AI tech. That's a much bigger debate best had elsewhere.
I will say that I have no interest in using AI art or writing for my projects. I've tried out a few programs to see what they can do, and sure, it can make some interesting looking stuff sometimes, but I would never consider anything generated by it to be something I can take credit for or pride in making. I can already write for myself, and while I am an extreme amateur at art, drawing and painting are skills I want to build up doing with my own hands, something AI generation cannot replace. Even using 3D programs, there is still manual effort applied in doing the character posing, scene setting, lighting, and "camera" work, so you're doing a lot more than just clicking a few buttons and letting things auto-generate. Likewise, I also do not feel the need to use AI as a crutch for a lack of creativity, and would feel no sense of accomplishment just asking the computer to make up stories and art for me based on some simple prompts.
On the other hand, and I am willing to concede some hypocrisy on my part here, I kinda feel less bad about utilizing text-to-speech technology, at least in the case of something like this, where I'm just toodling around with personal projects that I have no intention of trying to profit from. TTS tech has existed for decades already, and unlike with art and writing, you aren't asking the computer to just make something up for you. You still have to write the whole story or script, you still need to do all the audio editing that follows. The computer is just reading off what you wrote. Moreover, with newer speech-to-speech technology, you're also still doing all the actual voice acting yourself, while the program is basically just putting a filter over your voice to adjust how it sounds. ElevenLabs claims to have sourced their voices from willing volunteers and from contracted and compensated voice actors, so they seem less sketchy than a lot of the AI art companies. So, yeah, fuck it, I'm willing to play around with the tech for little pocket projects like these.
Once again, sorry if the video quality isn't very good, Blogspot seems to really want to compress things. If I ever do actually go through with making a little movie, I will definitely host the video elsewhere, even if I just have to throw it on a Google Drive or something.
(Truthfully, not much is probably going to come of this stuff, so I wouldn't hold your breath. I'll probably burn out on doing a bunch little "proof of concept" clips over the weekend, and then not update again for a year. Still, one never knows when the muse will drop by again.)
