top of page


AI art has been around for years now, improving at quicker speeds than ever, though it's still not fully embraced by all as an art form.

One of the biggest misconceptions is that you just type a few words and a masterpiece pops out. Therefore anyone can do it. While it's true 'something' will generate, even if you only use one word, it doesn't mean it will be a good composition.

It takes practice, lots of creativity and overall understanding of the AI you're using to generate great images - especially for those who have developed a recognizable aesthetic.

Another issue that some take with AI art is the training data - did it take it from the internet freely - how did it get it to where it is? This debate will never be satisfied and of course it took images from the public domain, whether anyone would have agreed to it or not.

But that's already happened and there's no turning back now. The question is - should the source of the training data determine whether or not to explore an entirely new creative expression? Judging by the millions of images that have already been generated, seems the answer's in the numbers.

**The AI images on this website are all created by (me) Melissa Victoria for SpiritSTEM.

My aesthetic is cosmic - sci-fi - contemplative. I create imagined worlds, explore the nature of reality and design futuristic fashion.

The image tools I use are Midjourney and DALL- E 3. I've tried many other image generators, but prefer Midjourney above all and DALL E-3 as a secondary, because it's within my custom GPTs.

As with images, I've tried many different video platforms -- and like everyone else in AI art, I'm waiting for public access to Sora by OpenAI, which will probably knock it so far out of the park, nothing else will compare.

  • Learn how to make your own images and video by getting the free guide in my STORE, Top Ten Tips (AI)

  • Check out my new and vintage art by visiting AI Art Galleries in the PORTFOLIO

  • Visit my STORE for free high res images


bottom of page