The development of generative AI models like Dall-E and ChatGPT has led to a surge in AI-generated content on the web.

This reliance on massive data sets has resulted in the inclusion of poor quality data, negative stereotypes, and problematic content.

The problem is expected to worsen as these models encode and amplify social stereotypes.

Synthetic articles, misinformation, and even music are flooding the web.

There is a risk of a recursive loop where AI models are trained using only synthetic data, perpetuating historical and societal inequities.

The implications of this for sectors like medicine, therapy, education, and law are yet to be fully understood.

The generative AI explosion of content may become a toxic dump with disastrous consequences.