Scientists Develop New AI on the Dark Web
OpenAI’s Large Language Models (LLMs), which are trained using a wide range of datasets, pull information from the dustiest corners on the internet.
What if, instead, such a model crawled through the dark Web — the seedy underbelly of the internet where you can host sites without your identity becoming public or available to law-enforcement? A group of South Korean researchers created an AI model called DarkBERT that indexes some of the most sketchy domains on the Internet.
This is a fascinating look at some of the most shady corners of the World Wide Web. These areas are synonymous with criminal and malicious activity, from the distribution of leaked information to the sale hard drugs.
Source:
https://futurism.com/the-byte/ai-trained-dark-web