askgaming

I'm not a fan of using gen ai in prototyping because even if you eventually replace the all the ai assets, the initial designs will inevitably seed and steer the artistic process, degrading the quality of the end result. A lot of the time when generating ai images or music, the prompter ends up having to settle for whatever the model decides to spit out, and their own artistic input gets lost along the way.

Also making placeholder assets for a game isn't even that hard? Even if it's something you scribble out in ms paint for 30 seconds, you're planning to fix it later. Plus creating your assets from scratch makes it much easier to iterate on the design yourself.

If it's being used for some sort of interactable gimmick, that's neat I suppose, but then you'll have the issue of requiring an internet connection and API access, meaning the game will stop functioning in the future.

And all of this is ignoring the ethical and environmental concerns of gen AI.