Large Language Models now generate over 100 million images daily using scraped training data.
Since the launch of ChatGPT in 2022, artificial intelligence has shifted from text generation to large-scale visual production. Data shows more than 30 billion AI images have been generated online, with 100 million created daily via ChatGPT alone. These systems, known as Large Language Models, utilize trillions of parameters to predict sequences and mimic human output. Technical foundations including neural networks and deep learning architectures allow these machines to produce still and moving images. However, significant ethical concerns persist regarding the training data used by these systems. Reports indicate this information was scraped from authors, photographers, and videographers without consent or compensation. As these models expand to serve as virtual companions and professional tools, real photographs are increasingly pushed aside. This transition marks a shift from the 1966 ELIZA chatbot to modern systems that many experts believe have finally passed the Turing Test.
Creative professionals face direct competition from systems trained on their own uncompensated work. The scale of automated image production threatens the market for authentic visual records and professional photography.
LoopEditions, our marketplace for buying original art, music and photography direct from independent creators, is at www.loopeditions.com. A machine-readable summary of this site is at /llms.txt.
This page needs JavaScript for the full experience. Read more about CreativLoop at www.creativloop.com.