About this Event
The rise of advanced Large Language Models (LLMs) like ChatGPT has significantly complicated the task of detecting AI-generated content. As these models become more sophisticated, bad actors are leveraging them to create highly convincing fake text, imagery, and audio, blurring the lines between genuine and artificial content. For large enterprises, this surge in high-quality, deceptive material undermines trust and can damage brand reputation. Traditional detection methods have become increasingly ineffective, leaving companies vulnerable to these sophisticated threats.
Can Java help?
Yes!!
To combat this issue, Johns Hopkins University Applied Physics Laboratory (JHUAPL) has developed a novel technique that pits a language model’s token embedding system against itself. This groundbreaking method is highly effective in identifying deep fakes and has successfully traced the source of a recent deep fake political robocall. In this presentation, we will dive into the mechanics of this technique using JHUAPL’s open-source tool Trinity, built with Java and JavaFX 3D. Attendees will gain insights into the practical application of this tool in detecting AI-generated content.
We may also play 3D asteroids!
Speaker: Sean M. Phillips is a senior software engineer at the Johns Hopkins University Applied Physics Laboratory who specializes in custom data visualization. Sean is a Java Champion and multiple Duke's Choice Award winner, providing research and capabilities on several domains, including Cislunar space defense, Brain-computer interfaces, and advanced cyber-physical effects.
Event Venue & Nearby Stays
BNY Mellon, 240 Greenwich Street, New York, United States
USD 0.00