Stable Diffusion has been making a massive name for itself lately. In more recent news, Lensa has used it to make magical avatars for celebrities and normal people like us. However, you may have missed the news a few months ago of a gaming fan upgrading game assets with it. It looks like the img2img AI technology may be the next trending topic.
Many of us gamers hail from the 80s/90s, and yes, I’m mostly speaking about myself. How many of you remember the DOS games with those really basic graphics? Well, frigis9 shared some images a few months ago on Reddit about how he used img2img technology with Visions of Chaos to upgrade the visuals with prompts.
If you’re thinking of doing the same, you may want to wait until the technology has a better way of understanding prompts. It’s quite challenging for the AI tech to understand what you want it to do. To that end, it takes plenty of trial and error before you receive an image that’s worth sharing.
Still, the images shared by frigis9 are quite good. The AI tech took the original images supplied to it, with some prompts for the output, and delivered upgraded visuals. We’re sure the human developers of the original games could do better, but it shows what the technology can do at this point. Of course, it should be more powerful by next year.
Yet again, we have an issue of copyright here. There are assets that are copyrighted to the companies that developed the games. While it’s fun making these portraits, designers using this AI technology won’t be able to sell them or make NFTs of them without the publisher’s permission. Or can they?
