Edmon Begoli on adversarial AI and its future implications.
Subscribe: Apple • Android • Spotify • Stitcher • Google • AntennaPod • RSS.
Edmon Begoli, leads the AI Systems R&D section at Oak Ridge National Laboratory (ORNL), where he is also a distinguished member of the ORNL research staff. Our conversation centered on his upcoming presentation at the Data+AI Summit, where he will describe the four principal categories of Adversarial AI and their future implications.
Highlights in the video version:
- Introduction to Edmon Begoli
What’s the mission of Oak Ridge National Laboratory?
Adversarial ML/AI and deep fakes
Access to tools for fine tuning machine learning models
Accuracy and API based access
Multi-modal models will advance the state of AI
Graph Neural Networks
Adversarial ML and Cybersecurity
Data poisoning
ML in industrial sector and protect AI against exploitation
Reinforcement Learning
Simulation Platform
Large scale distributed online learning
End to end learning in a large parameter space and their vulnerabilities
Related content:
- A video version of this conversation is available on our YouTube channel.
- Get Ready For Confidential Computing
- Jack Clark: The 2022 AI Index
- Nic Hohn and Max Pumperla: Reinforcement Learning in Real-World Applications
- Ram Shankar: Securing machine learning applications
- Nicholas Boucher: Imperceptible NLP Attacks
- Edmon Begoli: Hyperscaling natural language processing
FREE report:
[Image from Piqsels.]