How to fine-tune an ML model to create TV shows and movies with new titles, plot summaries, and scripts — I showed in a previous post how I fine-tuned GPT-J to generate Haikus, and the results were pretty good. For my latest experiment, GreenLIT, I wanted to push the limits of using GPT-J for creative writing to see if it could produce scripts for brand new TV shows and movies. …

How to use GLIDE and BSRGAN to create ultra-high-resolution digital paintings with fine details — I have been experimenting and writing about using AI/ML to create art from text descriptions for over a year now. During this time, I have noticed a significant increase in interest in this area, in part due to the burgeoning NFT art market. After looking at dozens of ML models…

How to generate rhythmic prose after fine-tuning a large transformer with phonemes — In this article, I’ll show you how I fine-tuned an AI system called GPT-J to create new Haikus, short poems following a form that originated in Japan. The key was to get my model, Deep Haiku, to see and understand the number of syllables in lines of poetry. I usually…

A discussion with Robert A. Gonsalves and Don Logan — The following is a conversation between photographer and digital artist Don Logan and me on the recent ruling by the US Copyright Office that AI-generated art cannot be copywritten in certain circumstances. The image in question can be seen above. The conversation continues on to explore related topics like the…

Using Machine Learning to change the gender, ethnicity, and age of peoples’ faces — Did you ever see an image that made you challenge some of your core beliefs? This happened to me the first time I saw a “south-up” map of the world like this.

How to use StyleGAN2, VQGAN, and GPT-3 to synthesize diverse characters from open-source images — I have previously written on Medium about using AI to create visual art, like abstract paintings and landscapes. During my research, I noticed that Generative Adversarial Networks (GANs) seem to have a hard time creating images of people. I decided to take this challenge [ahem] head-on, so I trained two…

Using GPT-J to analyze the lyrics of Rock & Roll songs — Question: Which topic is the most common in rock and roll lyrics? Answer: The Beatles summed it up in their song, “All You Need is Love.” As an experiment in topic modeling, I used the latest AI systems to analyze 12,000 songs by 50 rock bands. This article will walk…

How to use StyleGAN 2, VQGAN, and CLIP to create spooky images and videos for Halloween — There is a growing trend to use text phrases to direct Generative Adversarial Networks (GANs) to generate new images. When OpenAI introduced their CLIP [1] system for comparing text and pictures, it was a natural choice to use for text-to-image generation. Two popular GANs are StyleGAN 2 from Nvidia [2]…

Using the CLIP and VQGAN models to generate ChromaScapes, high-quality digital paintings for sale as NFTs on OpenSea — I first wrote about using Generative Adversarial Networks (GANs) to create visual art in August of 2020. For that project, MachineRay, I trained NVidia’s StyleGAN2 [1] with abstract paintings in the public domain to create new works. Since then, I have written a few more articles on using GANs to…